1

MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND

HOLD USERS’ ATTENTION?

Made You Look: Do Video Thumbnails with Portraits Attract and Hold Users’ Attention?

Megan K. Mendenhall

University of North Carolina at Chapel Hill

MATC Thesis

Laura Ruel, Chair

Chad Heartwood, Committee Member

Kathryn Kennedy, Committee Member

December 02, 2021

 2
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Table of Contents

Abstract............................................................................................................................................3

Acknowledgments............................................................................................................................4

Introduction......................................................................................................................................5

Review of the Literature..................................................................................................................7

Research Questions........................................................................................................................23

Procedures......................................................................................................................................24

Analysis…......................................................................................................................................29

Findings.........................................................................................................................................31

Implications...…………………………………………………………………………..………...41

Limitations…………………………………………………………………………………….…42

Conclusions………………………...………………………………………………………….....44

Further Research…………………………………………………………………………………45

Appendix........................................................................................................................................46

References....................................................................................................................................106
 3
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Abstract

This qualitative pilot study investigates if video thumbnail designs with portraits attract

users’ attention faster and hold their attention longer than thumbnail designs without portraits

during information-seeking tasks. The study also investigates if users perceive content

represented by thumbnail designs with portraits as more trustworthy than content represented by

designs without portraits. Additionally, the study asks if, during information-seeking tasks, users

show a bias toward engaging with content represented by designs with portraits over designs

without portraits.

The study used pre-and post-surveys, eye-tracking data, and retro-active think-a-loud

interviews to examine these questions. The findings suggest that designs with portraits do not

draw users’ attention faster than designs without portraits. However, designs with portraits hold

users’ attention longer than designs without portraits only in some information-seeking tasks; for

other tasks, designs without portraits hold users’ attention longer than those with portraits.

Findings also suggest that some users rely on thumbnail designs as a secondary source of

information when assessing the trustworthiness of the video content. Additionally, the findings

suggest that during search scenarios when users value information from a source belonging to a

specific group or identity, users perceived designs with portraits of sources belonging to the

desired group or identity as more trustworthy than thumbnail designs without portraits or designs

with portraits of sources from other groups or identities.

The findings also suggest that users are not bias toward engaging with content

represented by thumbnail designs with portraits over designs without portraits during

information-seeking tasks. Furthermore, the findings suggest that users use thumbnail designs as

secondary sources of information to assist with determining the style or tone of video content.
 4
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Acknowledgments

I want to thank the following people who encouraged and supported me throughout my graduate

education.

To Laura Ruel, I love your enthusiasm and encouragement. Thank you for guiding my seed of

an idea into a completed project and constantly inspiring me to approach the world with "how

might we" questions.

To Chad Heartwood and Kathryn Kennedy, thank you to my highly patient committee

members who worked with me as my timeline for completing my thesis project expanded from

six months to a year and a half.

To the MATC Cohorts of 2016, 2017, and 2018, thank you all for being there with sanity

checks, venting sessions, and words of encouragement.

To Jenny Cook, Denise Haviland, and Camille Jackson, I could not have made it through Strat

Comm without all of you cheering me on.

To Emilie Poplett, thank you for allowing me to use your business, Homegrown Videography,

as a testing ground while I learned how to design a mobile app.

To Ashley Wolf, thank you for your insight and for sharing your thumbnail designs. This project

could not have happened without you.

To Emily McGinty and the Duke Campus Farm, thank you for graciously hosting me during a

pandemic so that I could write my proposal in a much-needed quiet place.

To my mom, Kathleen Longbrake, thank you for the countless pep talks and sanity checks!

To my husband, Grayson Mendenhall, thank you for being a single parent to Karl, Caedon, and

Elena on countless nights and weekends, believing in me when I wasn't sure I believed in myself

and for loving me when I was super stressed out and not so easy to love.
 5
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Introduction

Users currently watch more than 500 million video videos on YouTube each day. Zenith

media forecasts this trend to continue and predicts that by 2021, the average person will spend

100 minutes a day watching online video (zenithmedia.com, 2019). Marketers are responding by

planning to incorporate more video content in their marketing strategies (Khabab, 2020).

According to a 2020 Marketer and Consumer Survey by MotionCue, 86 percent of marketers’

plan to incorporate more videos into their future marketing strategies (Motioncue, n.d.).

According to the same survey, 65 percent of consumers cite YouTube as their preferred platform

when watching video content (Motioncue, n.d.).

Many consumers turn to YouTube, a video hosting site and social networking platform,

to find information, making it the second-largest search engine behind Google (Davies, 2018;

Patel, 2016; Wagner, 2017). While there is some debate surrounding YouTube's status as a

major search engine versus an internal search engine (Rainwater, 2013), YouTube processes

more than three billion searches a month. Its search volume is larger than that of Bing, Yahoo,

AOL, and Ask.com combined (Wagner, 2017), proving that it is a widely popular and highly

visited site by users seeking information about a wide range of topics (Spangler, 2019).

YouTube presents searchers with a list of videos represented as thumbnails that provide

users a quick visual snapshot of the videos' content. Video thumbnails are important because

they serve as a first impression for a brand (Brown, 2017) and play a critical role in increasing

click rates and view counts of videos, which in turn can lead to increased conversion rates

(Agbbon, 2019).

To attract users' attention, marketing experts recommend incorporating images of human

faces into video thumbnails (Shiffman, 2015). Research supports the idea that within social
 6
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

networking communities, users are more likely to engage with content that features human faces

than with content that does not (Bakhshi et al., 2014).

While there is little research into how users interact with video thumbnail designs within

the context of information-seeking tasks, research does suggest that human faces attract attention

over other objects (Hershler & Hochstein, 2005). This study aims to test the hypothesis that

when users engage in an information-seeking task for video content, they will show a stronger

bias toward video thumbnail designs that include visuals of faces than toward designs without

faces.

The first section of the literature review discusses the top-down and bottom-up theories

of visual attention and capture. The next section examines information-seeking behavior, user

mindsets during searches, and user scanning patterns of search engine results pages. The final

section discusses users' attention toward and engagement with visual content featuring human

faces.
 7
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Review of the Literature

Top-Down and Bottom-Up Theories of Visual Attention

Alfred Yarbus, a pioneer in the study of eye movements, found that observers' eye

movement patterns when free viewing a scene differed from their eye movement patterns while

viewing the same scene during a target search task (Yarbus, 1967). In addition, Yarbus found

search task criteria influenced eye-movement patterns. Since eye movements are the output of

visual attention, scholars concluded that in some instances, cognitive influences direct observers'

scanning patterns (DeAngelus & Pelz, 2009; Tatler, 2009; Yarbus, 1967). These situations are

referred to top-down visual processing because the observers' visual attention is direct by the

need to complete a cognitive task. Before this observation, scholars believed that visual attention

resulted from bottom-up visual processing, where the most salient objects in a visual scene

automatically drew observers' visual attention.

The implications of Yarbus' finding has caused debate among cognitive psychologists as

to how visual attention is processed. Folk et al. examined the conditions under which bottom-up

processing could be induced during a search task and the properties of stimuli that could trigger

exogenous controls of attention (Folk et al., 1992). They found that bottom-up processing only

occurs when distracting stimuli have similar properties to target stimuli (Folk et al., 1992).

Because they found that bottom-up processing did not take place in conditions when the

distracting stimuli did not share characteristics of the target stimuli, Folk et al. concluded that

visual attention is dependent on the attentional control settings dictated by the search task and

that top-down processing does play a role in visual attention and capture (Folk et al., 1992).

However, other studies have found conflicting results with these findings. Theeuwes

argues that during parallel search visual tasks, it is impossible for the visual processing to
 8
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

provide top-down guidance to the point where visual attention can be task-driven and determined

that visual attention and capture is determined solely by stimulus salience (Theeuwes, 1991). But

then in a later study, Theeuwes found that task-driven attentional sets did play a role in the later

stages of visual capture, but not in the initial pre-attentive stages of visual attention (Theeuwes,

2010). While the findings of Folk et al. and Theeuwes are not in agreement, Theeuwes's later

results are more in line with those of Folk et al., in showing that top-down processing influences

visual attention in later stages in the attention process (Theeuwes, 2010).

Henderson and Hayes found that in photos of real-world scenes, meaning played a more

influential role than salience in guiding visual attention (Henderson & Hayes, 2017). In their

study, they created meaning maps by representing the spatial distribution of semantic

information in scenes. Then they made salience maps based on image features found in the same

scenes. Using heatmaps from eye-tracking experiments, they detected which areas of the scenes

drew the most visual attention from participants and mapped these areas of greater visual

attention to either location of objects with high meaning or objects with high visual salience.

They found that objects with high meaning attracted more visual attention than high salience

(Henderson & Hayes, 2017). While what is meaningful to participants may differ from the target

of a visual search task, attending to either stimulus shows that participants must exert some

cognitive control over where they direct their visual attention. And this observation implies that

top-down processes play a role in visual attention, even in conditions when specific visual search

targets are not present to cue participants (Henderson & Hayes, 2017).

Other studies involving visual attention have found that changes to attentional capture

can result from changes in participants' naturally occurring goal states (Cunningham & Egeth,

2018). Cunningham and Egeth found that participants' attentional capture is distracted by task-
 9
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

irrelevant images under certain conditions because specific visual stimuli can trigger a change in

users' task goals (Cunningham & Egeth, 2018).

In their initial experiment, Cunningham and Egeth found that participants were more

distracted by task-irrelevant images of energy-dense food (candy) than by visuals of non-food

objects and low-energy foods. However, in a second experiment, where participants ate a small

amount of candy before participating in the experiment, the candy pictures' distracting effect was

significantly less than those in the first condition who did not consume candy before the

experiment (Cunningham & Egeth, 2018).

These findings demonstrate that task-driven attentional processing can be interrupted by

stimuli-driven processing in certain conditions if stimuli contain specific characteristics that

prompt a change in participants' goal-states (Cunningham & Egeth, 2018). It also showed that

naturally occurring goal states' changes are difficult for users to ignore, but they are highly

flexible (Cunningham & Egeth, 2018). While the study specifically examined how images of

energy-dense food can cause changes in participants' goal state, it is unclear if other visual

stimuli would have the same disrupting effect on participants' visual attention. However, these

findings imply that certain visual stimuli, specifically ones that can trigger particular in born

preferences, may cause changes in participant’s goal states, resulting in their attention being

shifted toward task-irrelevant visuals (Cunningham & Egeth, 2018).

The research collectively indicates that visual attention and capture are influenced by

task-driven (DeAngelus & Pelz, 2009; Tatler, 2009; Yarbus, 1967) and stimulus-driven

(Theeuwes, 1991) processes. And that in the absence of specific search task requirements,

participants are likely to show an attentional bias toward areas or objects within visual scenes

that hold meaning to them over highly salient stimuli (Henderson & Hayes, 2017). It also
 10
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

indicates that task-irrelevant stimuli, when viewed under certain conditions, could trigger

changes to users' goal-tasks and cause users to direct their attention toward task-irrelevant visuals

(Cunningham & Egeth, 2018).

There are many implications of this research for the design of video thumbnails. Top-

down processing will play a role in directing users’ visual attention because they are likely to

view thumbnails while engaged in information-seeking tasks. However, because some

information searches can be vague (Wilson, 1999) users will not always have a clearly defined

search target in mind. Therefore, it is expected that users will be directed toward visuals that

hold meaning to their search (Henderson & Hayes, 2017). Additionally, some visuals are likely

to distract users' attention, specifically those that trigger inborn preferences, such as the desire for

energy-dense food (Cunningham & Egeth, 2018). Effective video thumbnails will incorporate

these findings into their designs and seize opportunities to integrate relevant and attention-getting

visuals into their designs.

User Information-Seeking Behavior

When users perform a search in a search engine, they are engaging in an information-

seeking activity (Wilson, 1999). T.D. Wilson conducted a study that compared five models of

information seeking and searching behavior. Because these models overlapped in their scope and

addressed similar concepts and behaviors, his goal was to create a new model that incorporated

and nested related facets of the pre-existing models (Wilson, 1999). Some of the similarities that

he found among the models are that users seek information with various approaches, sometimes

without a clearly defined target, and in some cases will refine search criteria as more information

is gathered (Wilson, 1999). While these concepts may seem obvious, it is crucial to understand
 11
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

that searches are not always binary, and users may not know what they are seeking at the start of

their search process (Li & Belkin, 2008; Wilson, 1999).

User Mindsets Effects Perception of Search Success

Research into how users evaluate music searches reveals that they judged their searches

on two criteria: success and effort (Hosey et al., 2019). Hosey et al. also found that users'

mindsets played a role in how they evaluated the success of their music searches and their

tolerance for effort in search tasks (Hosey et al., 2019). The study found that most participants

searched for music with a focused mindset and considered a search successful if they found the

specific items they were seeking and showed a low tolerance for effort (Hosey et al., 2019).

Some participants approached the search from an open mindset, searching for content with a

particular theme, but without a specific target. These participants considered a search successful

if they found items that closely matched their search criteria and showed some tolerance for

search effort (Hosey et al., 2019). Only a few participants approached the search from an

exploratory mindset. These participants had difficulty assessing their searches' success because

their goals were more aligned with learning about a specific area than on finding a particular

item (Hosey et al., 2019).

These findings demonstrate that users' mindsets play a role in defining a successful

search and how much effort they will tolerate completing a search (Hosey et al., 2019). Video

thumbnail designs should consider these mindsets and incorporate visuals easily recognizable to

users with varying mindsets.

User Scanning Patterns of SERPs

A recent report from the Nielsen Norman Group found that user behavior on search-

engine-results-pages or SERPs has evolved to better take advantage of current, more complex
 12
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

layouts of SERPs (Moran & Goray, 2019). Previous studies showed that users scanned SERP

results sequentially, and their scan paths moved down the page in an F-shaped pattern (Pernice,

2017). Now users scan SERPs in a pinball pattern where their scan paths bounce back and forth

throughout the SERP, much like the pinball bounces throughout a pinball machine (Moran &

Goray, 2019).

This finding implies that users are less likely to automatically engage with the first item

on the SERP, thereby partially eliminating the location advantage once enjoyed by content that

appeared in the upper left section of SERPs (Moran & Goray, 2019). Simultaneously, pinball

scanning patterns lead to a more even distribution of user attention across the SERP, and

competition for user attention among individual thumbnails increases (Moran & Goray, 2019;

Royse, 2020). Thumbnail designs should be optimized to stand out when viewed in the highly

competitive environments of SERPs (Royse, 2020).

Lewandowski and Kammerer reviewed 41 papers in which eye-tracking was used to

examine the effects of a specific factor on users' viewing behavior of SERPs (Lewandowski &

Kammerer, 2021). However, in their review, Lewandowski and Kammerer cited only one study

that examined how images might play a role in how users scan SERPs (Lewandowski &

Kammerer, 2021).

The study examined if pupil size could be used to discriminate users' perceived relevance

of text and images on SERPs (Oliveira et al., 2009). The initial experiment found that when users

were exposed to relevant text SERP results, their pupil dilation would increase relative to

irrelevant text results (Oliveira et al., 2009). However, no significant difference was found when

users were exposed to relevant image results relative to irrelevant image results (Oliveira et al.,

2009). Therefore, in the second experiment, Oliveira et al. simplified the discrimination task.
 13
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

After modifying the discrimination task, they found that when users were exposed to relevant

image results, their pupil dilation would increase relative to irrelevant text results (Oliveira et al.,

2009). Thus, Oliveira et al.'s preliminary findings suggest that pupil dilation can be used to distill

information about the relevance of both text and images SERP results (Oliveira et al., 2009).

However, with more ambiguous discrimination tasks, there may not be significant differences in

pupil dilation when exposed to relevant images relative to irrelevant images (Oliveira et al.,

2009).

Collectively, the literature shows that users seek information with a variety of

approaches, sometimes without a clearly defined target and in some cases will refine search

criteria as more information is gathered (Li & Belkin, 2008; Wilson, 1999). Users' mindsets also

affect how they approach searches, how much effort is put into searches, and how the outcomes

of searches are perceived (Hosey et al., 2019). Also, users are likely to scan SERPs in a non-

linear way (Moran & Goray, 2019). Finally, there is very little research on how specific factors

such as images, effect users’ viewing behavior of SERPs (Lewandowski & Kammerer, 2021). As

such, designers should take users’ information-seeking behaviors, mindsets, and SERP scanning

patterns into consideration when optimizing thumbnail designs.

Human Faces and User Attention

Scholars offer conflicting theories on how visual attention and capture are processed

(Folk et al., 1992; Henderson & Hayes, 2017; Theeuwes, 2010). However, additional research

indicates that the presence of a human face, regardless of its relevance to users' task goals, will

attract users' visual attention toward the face and away from task-relevant visuals (Ariga &

Arihara, 2017; Hershler & Hochstein, 2005; Sato & Kawahara, 2015). This finding implies that

human faces automatically attract users' attention (Hershler & Hochstein, 2005) and disrupt top-
 14
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

down attention sets away from goal-related visuals (Sato & Kawahara, 2015). Hershler and

Hochstein found that participants did not show the same bias toward animals' faces as they did

toward human faces, and Ariga and Arihara's research found that pareidolia faces do not attract

participants' attention in the same way as actual human faces.

The fusiform facial area (or FFA) is part of the human visual system that is sensitive to

human faces (KLI, 2016; Weinschenk, 2019), responsible for facial recognition (Kanwisher et

al., 1997), and for processing facial expressions of emotion (Weinschenk, 2019). The FFA

activates when we see a face either straight or in profile (Weinschenk, 2019) and this activation

helps to explain why faces are unique in automatically commanding users' attention.

Heatmaps from eye-tracking studies of visual designs that included human faces show

that users pay more attention to faces than to non-face objects (KLI, 2016; Weinschenk, 2019).

In designs where a face is looking straight on, users will direct their attention to the face's eyes,

as if making eye contact (KLI, 2016; Weinschenk, 2019). In designs where a face is looking off-

camera or toward an object, users will mimic the face's gaze path by looking in the same

direction as the face (KLI, 2016; Weinschenk, 2019).

The literature shows that humans are hard-wired to pay attention to faces (Hershler &

Hochstein, 2005). However, not every face is equally effective in attracting and holding users'

attention (Aharon et al., 2001; Chudley, 2013; Shimojo et al., 2003). This section will discuss

how faces' perceived level of attractiveness affect users' reactions toward them, how faces affect

users' level of trust toward a digital product, and how faces affect users' task performance.

Attractive and Unattractive Faces Have Differing Effects on User Attention

While research has shown that human faces effectively capture users' attention (Sato &

Kawahara, 2015) not all human faces are equally effective in sustaining users' attention (Valuch
 15
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

et al., 2015). Scholars agree that users spend more time looking at faces they find attractive and

less time viewing faces they deem unattractive (Aharon et al., 2001; Shimojo et al., 2003; Sui &

Liu, 2009), and in some cases, ignore unattractive faces after their initial gaze (Chudley, 2013).

Valuch et al. 's eye-tracking studies also found differences in how male and female users' attend

to attractive faces: males spend more time viewing and show a stronger bias to view attractive,

opposite-sex faces than female users (Valuch et al., 2015). It is worth noting that the literature

does not refer to non-binary individuals. It is unclear if the current research referring to male and

female-identifying individuals applies to non-binary identifying individuals. Additionally, the

literature does not address non-heteronormative individuals and how or if they showed a stronger

bias toward opposite-sex faces or same-sex faces.

While users can rate a face as attractive or unattractive, it is difficult to name the specific

features that determine this attraction. Studies indicate that facial attractiveness judgments are

made holistically, rather than dependent on a particular facial characteristic (Hershler &

Hochstein, 2005). However, general facial features, such as symmetry and averageness, have

been shown to influence users' attractiveness judgments of faces (Grammer & Thornhill, 1994;

Thornhill & Gangestad, 1993). Thornhill and Gangestad compared composite portraits made up

of multiple faces with faces of individuals. The composite faces were more symmetrical than the

individual faces and had facial dimensions closer to average measurements than the individual

faces. Participants rated the composite faces as more attractive than the individual faces. They

argue that the more symmetrical and average faces may be preferred to less-symmetrical and

less-average faces because owners of average faces possess a more diverse set of genes, resulting

in more successful mating (Thornhill & Gangestad, 1993) .

 16
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Studies have found that facial characteristics and skin tone play a role in judgments of

facial attractiveness (Burke et al., 2013; Lewis, 2010, 2011, 2012; Rhodes et al., 2001, 2005;

Stepanova & Strube, 2018). However, the studies have produced somewhat incongruent

findings. For example, Burke et al.’s study compared attractiveness ratings given by participants

of differing origins and ethnicities to faces with Eurocentric features and skin tones, East Asian-

centric features and skin tones, and Afrocentric features and skin tones (Burke et al., 2013). The

study found that both female and male participants rated opposite-sex faces with Eurocentric

features and skin tones as more attractive than faces with Asian-centric and Afrocentric features

and skin tones (Burke et al., 2013). Burke et al.’s study also found that female participants rated

individual male faces with Afrocentric features and skin tones as equally attractive to individual

male faces with Asian-centric features and skin tones. Additionally, Burke et al. found that

female participants rated compound male faces with Afrocentric features as slightly more

attractive than compound male faces with Asian-centric features (Burke et al., 2013).

However, Burke et al.’s findings conflict with other scholars. For example, Lewis (2011)

found that both British (Eurocentric) males and females rated Afrocentric male faces as more

attractive than male faces with Eurocentric or mixed-race features and skin tones (Lewis, 2011).

Additionally, Lewis (2012) found that British women of different racial groups rated males with

Afrocentric facial features and skin tones as more attractive than males with Eurocentric or

Asian-centric facial features and skin tones (Lewis, 2012). Burke et al. explains this

inconsistency by arguing that perceptions of masculinity and femininity drive assessment of

attractiveness, and males with Afrocentric facial features and skin tone could be perceived as

more, or at least as, masculine as male faces with Eurocentric, Asian-centric or mixed race

features and skin tone (Burke et al., 2013).

 17
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Other scholars have found that mixed-race faces are perceived as the most attractive

compared to faces with single race features and skin tones (Lewis, 2010; Rhodes et al., 2005;

Stepanova & Strube, 2018). For example, Lewis (2010) found that Eurocentric (white)

participants rated both male and female mixed-race facial features and skin tones as the most

attractive when compared with single-race Afrocentric or Eurocentric faces (Lewis, 2010).

Likewise, Rhodes (2005) found a preference for mix-race faces over single-race faces. Her 2005

study showed that both male and female participants rated opposite sex composite faces with

mixed-race features and skin tones as the most attractive when compared to Eurocentric faces

(white) and Asian-Centric (Japanese) faces (Rhodes et al., 2005). Additionally, Stepanova and

Strube (2018) found that female raters of various races preferred male faces with mixed-race

features and skin tones over male faces with solely Afrocentric and Eurocentric features

(Stepanova & Strube, 2018).

Lewis theorizes that heterosis many play a role in the preference for mixed-race faces

over single race faces (Lewis, 2010, 2011). Heterosis occurs when the offspring of two

genetically different parents have traits that are enhanced or superior to their parents due to

varied genetic contributions from each parent. Biologically speaking, these hybrid offspring tend

to be more genetically fit than offspring from parents with more similar genetic contributions

(“Heterosis,” 2021). Lewis posits that individuals with mixed-race facial features make be

perceived as more attractive than individuals with single-race facial features because mixed race

features could be an indication of genetic diversity and fitness (Lewis, 2010, 2011). Another

theory is that faces with mixed-race features are more likely to have less exaggerated and more

general facial features than single-race faces and that symmetry and averageness play a role in

the perception of attractiveness (Grammer & Thornhill, 1994; Thornhill & Gangestad, 1993).
 18
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

While several studies have demonstrated a preference for mixed-race faces, there are

studies where mixed-races faces were rated as equal to or as no more attractive than single race

faces (Burke et al., 2013; Rhodes et al., 2001) For example, Rhodes et al (2001) found that

Asian-Centric (Chinese) participants rated Eurocentric (white) faces, Asian-Centric (Chinese)

and mixed-race faces as equal in attractiveness (Rhodes et al., 2001). Similarly, Burke et al.

found that male and female participants rated mixed-race faces as no more attractive than single-

race faces (Burke et al., 2013).

While the details of how an individual's facial features and skin tone directly relate to

others' perception of their attractiveness remain unclear, the literature suggests both play a

complex and complicated role.

Scholars have also investigated if the physical traits of individuals affect judgments of

facial attractiveness (Laeng et al., 2007; Valuch et al., 2015). A 2007 study found that blue-eyed

males (participant trait) displayed a stronger bias toward blue-eyed female faces (stimulus

characteristic) than toward brown-eyed female faces (Laeng et al., 2007). Laeng et al. explain

this finding as a result of an evolutionary adaptive strategy of blue-eyed men to maximize their

"assurance of paternity" (Laeng et al., 2007). However, a 2015 study also investigated this

question and found that blue-eyed males did not display a stronger bias toward blue-eyed female

faces (Valuch et al., 2015). Valuch et al. suggest that a difference in sample size between the

two studies could be the reason for the inconsistent findings.

Collectively, the research suggests that more attractive faces hold users' attention longer

than less-attractive faces (Aharon et al., 2001; Shimojo et al., 2003; Sui & Liu, 2009) and

attractive, opposite-sex faces capture male users' attention, but not female users' attention more

quickly (Valuch et al., 2015). Research findings imply that faces are processed holistically
 19
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

(Hershler & Hochstein, 2005), and that symmetrical and average faces are generally rated as

more attractive than asymmetrical and faces with facial dimensions outside the norm (Grammer

& Thornhill, 1994; Thornhill & Gangestad, 1993). The literature suggests that an individual’s

facial features and skin tone play a complex and complicated a role in how others rate their

attractiveness (Burke et al., 2013; Lewis, 2010, 2011, 2012; Rhodes et al., 2001, 2005;

Stepanova & Strube, 2018). It is unclear if participants' traits play a role in users' judgment of

facial attractiveness (Laeng et al., 2007; Valuch et al., 2015).

Faces and Their Differing Effects on Users' Perception of Trust

There is little research discussing how the inclusion of human faces in video thumbnail

design affects users' perception of trust in the video, or if an increased perception of trust

correlates to increased user engagement. However, studies examining how the inclusion of staff

portraits in the designs of e-commerce websites affects users' perception of the website's

trustworthiness have mixed findings (Riegelsberger et al., 2003; Riegelsberger & Sasse, 2001;

Steinbrück et al., 2002; Tullis et al., 2009). Reigelsberger and Sasse conducted extensive studies

on how the inclusion of "elements of social interaction" affects users' perception of a website's

trustworthiness (Riegelsberger & Sasse, 2001). Staff portraits are considered "elements of social

interaction" in the context of online shopping because these portraits serve as digital stand-ins for

face-to-face interactions that likely occur in physical environments (Riegelsberger & Sasse,

2001). Adding elements of social cues to digital experiences is known as "virtual re-embedding"

because these digital elements are designed to re-embed users' trust in vendors during a digital

experience (Riegelsberger & Sasse, 2001).

Interestingly, their research found that users did not consciously engage with staff

portraits while performing mock online shopping tasks. Furthermore, Reigelsberger and Sasse
 20
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

found that "experienced and trusting internet users" viewed staff portraits negatively, citing that

the portraits did little to increase these users' already high trust levels (Riegelsberger & Sasse,

2001). However, Steinbrück et al. found that including portraits of customer representatives

increased users' perception of the site's trustworthiness, regardless of the users' Internet

experience and trust levels (Steinbrück et al., 2002). Steinbrück et al. do not explain why their

results differed from those of Reigelsberger and Sasse’s studies. However, the discrepancies in

findings could result from the differing scopes of the studies. Reigelsberger and Sasses' research

was more broadly focused on how virtual re-embedding effects users' perception of trust in e-

commerce websites (Riegelsberger & Sasse, 2001), while Steinbrück et al. examined how one

social cue (addition of staff portrait) affects users' perception of trust in a digital product

(Steinbrück et al., 2002). Overall, the research suggests that including portraits in digital

interfaces may not automatically result in an increase in users’ perception of trust interface.

Faces Effects on Users' Task Performance

While getting users' attention is critical for user engagement, there are particular contexts

where the presence of a face can distract users from task-related goals (Giang, n.d.; Tullis et al.,

2009). Tullis et al. hypothesized that including a portrait near task-relevant text would draw

users' attention to the information faster than if no portrait was present. To test this theory, they

created two versions of the same webpage. One version included a portrait near the task-relevant

information. The other version did not. Participants completed the same task on each version of

the page. The findings showed that participants in the portrait condition correctly completed the

task 78 percent of the time. In comparison, participants in the no portrait condition correctly

completed the task 93 percent of the time (Tullis et al., 2009). Also, the portrait condition

participants took longer on average to complete the task than those in no portrait condition
 21
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

(Tullis et al., 2009). Overall, participants in the portrait condition rated the overall ease of

finding information on the page lower than those in the no portrait condition (Tullis et al., 2009).

Also, of interest is that for the participants in the portrait condition, the eye-tracking analysis

showed they rarely fixated on the portrait (Tullis et al., 2009). The study suggests that in certain

task-oriented conditions, faces can negatively impact on user performance.

Tullis et al. suggest that their findings could result from users' overexposure to faces in

web ads and theorize that users may have learned to "tune out" faces (Tullis et al., 2009). It is

important to note that Tullis et al. used a portrait featuring a woman looking forward as if

making eye contact with users. They might have had different results if they had used a portrait

featuring a person gazing toward task-relevant information, as users typically follow the gaze

path of faces (KLI, 2016; Weinschenk, 2019).

Additional eye-tracking studies have found similar results on user performance. For

example, TheLadders.com, an online job search service, conducted an eye-tracking study of

recruiters' eye movements while reviewing candidates' resumes. A heat map from the study

showed that recruiters spent more time viewing candidates' portraits than more task-relevant

sections of the resumes, such as candidates' current positions, education, skills, and previous

work experience (Giang, n.d.)

While faces can attract users' attention, these studies provide evidence that faces can also

disrupt users' attention away from other more task-relevant information. These studies examined

user behavior on different platforms (website and resumes) and the user tasks varied between

studies. However, both studies demonstrated that faces could disrupt users' attention away from

task-relevant information. It is not clear if the presence of faces within video thumbnails' design

will have a similar task disrupting effect on users who are seeking information from SERPs.
 22
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

However, the context in which users view video thumbnail designs should be considered when

designing thumbnails that include portraits, as faces can disrupt some user tasks (Giang, n.d.;

Tullis et al., 2009).

 23
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Research Questions

The literature provides evidence that users’ visual attention is influenced by top-down visual

processing especially in situations when users are given specific cues on what and where to

direct their attention (DeAngelus & Pelz, 2009; Folk et al., 1992; Henderson & Hayes, 2017;

Tatler, 2009; Yarbus, 1967). However, research into information-seeking behavior reveals that

users often engage in a variety of non-linear approaches during exploratory searches, such as

refining query terms and readjusting task goals as they discover new information (Li & Belkin,

2008; Wilson, 1999). Because of the evolving nature of exploratory searches, defining a static

task goal is challenging. In the absence of a well-defined task goal or search target, will the

presence of faces attract users’ attention over other visuals? The study aims to answer the

following questions regarding users’ attention and reactions to video thumbnail designs

including portraits:

1. When users do information-seeking tasks, do video thumbnail designs with portraits

attract users’ attention faster than video thumbnail designs without portraits?

2. When users engage in information-seeking tasks, do video thumbnail designs with

portraits hold users’ attention longer than thumbnail designs without portraits?

3. Do users perceive videos with thumbnail designs that include portraits as more

trustworthy than videos with designs that do not have portraits?

4. Will users’ show a bias toward engaging with video content represented by thumbnails

with portraits than by thumbnails without portraits?

 24
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Procedures

The qualitative pilot study used an eye-tracking device to gather participants' eye

movements and fixations as they viewed video thumbnail designs within eight mock YouTube

SERPs (see Appendix A). Following the eye-tracking session, participants engaged in a

retrospective think-aloud exercise, where they watched a recording of their eye movements while

verbally describing their thought processes as they responded to the scenarios within each mock

SERP environment (Bojko, 2013; Retrospective Think-Aloud | Qualitative Research | Tobii Pro,

2019).

The eye-tracking and retrospective think-aloud data were collected from August 29,

2021, through September 4, 2021. The mock YouTube SERPs (Appendix A) are based on eight

search terms representing words and phrases corresponding to eight videos from Duke

University's video series, "Extra Credit: Life's Curiosities Explained." This video series features

experts providing advice and explanations for a wide range of topics which provides a diverse set

of mock SERPs. Additionally, the individual videos can be appropriately represented by

thumbnail designs featuring portraits and non-portrait visuals.

Creation of the Mock SERP Environments

Each of the mock SERP environments (Appendix A) were created using screenshots of

YouTube SERP returns taken on September 26, 2020. The YouTube SERP returns were

generated by using search terms that correspond to the titles and topics of eight videos from the

“Extra Credit” series (see Appendix B). Each mock SERP environment accommodated five

video thumbnails or query returns. During the experiment, one of three versions of the

corresponding video thumbnail design (see Appendices C through E), was placed into one of the

five query return positions within the corresponding mock SERP environment.
 25
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Placement of target thumbnail design within the mock SERP environments

Within each mock SERP, the query returns were assigned a number to indicate their

position within the environment. The query return in the top position in each mock SERP is

assigned number one, the next query return is assigned number two, the query return in the

middle position is assigned number three, the query return second from the bottom position is

assigned number four, and the query return in the bottom position is assigned number five.

The location of the target thumbnails within each mock SERP was determined by using a

random number generator, random.org (RANDOM.ORG - True Random Number Service, n.d.),

that randomly selected a number between one and five. The randomly selected number

corresponds to the position of the query return within the mock SERPs. The positions of the

target thumbnails were randomly selected to minimize the effects of the participants’ eye

movements becoming targeted as they progressed through the experiment (Bojko, 2013).

The positions of the target video thumbnails within each of the eight mock SERPs were

as follows: The target video thumbnail was assigned position five within the mock SERP for

higher education policy environment. The target thumbnail was assigned position four within the

mock SERP for too much homework environment. The target thumbnail was assigned position

one within the mock SERP for pop culture and Bible environment. The target thumbnail was

assigned position three within the mock SERP for Black comedy and activism environment. The

target thumbnail was assigned position five within the mock SERP for randomized roommates’

environment. The target thumbnail was assigned position two within the mock SERP for soda

tax environment. The target thumbnail was assigned position three within the mock SERP for is

stress bad environment. The target thumbnail was assigned position two within the mock SERP

for sugar cravings environment.

 26
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Participant Groups

Participants had normal to corrected vision, were fluent English speakers, ranged in age

from 18 to 65, and were recruited from residents living in the Research Triangle region of North

Carolina. (“Research Triangle,” 2021). The in-person eye-tracking experiments took place at the

three locations in Durham and Orange counties of North Carolina. Both Durham and Orange

counties are included in the Research Triangle region of North Carolina.

Usability research has shown that five people can find up to 80 percent of all usability

issues in digital multimedia experiences (Krug, 2014; Ruel, 2007). The goal of this study was to

determine if the use of portraits in video thumbnails designs is effective in attracting and

convincing users to engage with the video when compared to designs with non-portrait visuals

and text only designs. Therefore, the 15 participants in the study were randomly assigned to one

of three user testing groups, and each group had five participants.

Each group viewed one of the three versions of thumbnail designs for the eight videos

from the “Extra Credit” series. The thumbnail versions of each video were displayed within the

corresponding mock SERPs environments (see Appendix A). Participants in group one viewed a

set of video thumbnail designs featuring a portrait of a person featured in the corresponding

videos from the “Extra Credit” series (see Appendix C). During the study, participants in group

one viewed the portrait thumbnail designs within their corresponding mock SERP environments

(see Appendix F).

Participants in group two viewed a set of video thumbnail designs with relevant, but non-

portrait visuals (see Appendix D). The non-portrait visuals were selected using the search terms

corresponding to individual videos’ titles and topics (see Appendix B). During the study,
 27
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

participants in group two viewed the non-portrait thumbnail designs within their corresponding

mock SERP environments (see Appendix G).

Participants in group three viewed video thumbnail designs with text only (see Appendix

E). During the study, participants in group three viewed the text-only video thumbnail designs

within their corresponding mock SERP environments (see Appendix H). Background colors and

text treatment remained the same across all three versions of the thumbnail design.

Pre-Study Survey

Participants in all groups took a pre-study survey (see Appendix I) before viewing each

mock SERP. The pre-study survey (see Appendix I) collected participants’ self-reported

knowledge and expertise of the search topic and level of interest in the search topic.

Eye-Tracking Data Collection

The study collected eye-tracking data from all 15 participants as they view the eight mock

SERP environments. Participants’ eye movements as they looked over each of the eight mock

SERPs were recorded and displayed as gaze plots (see Appendices M through T). Along with

collecting the participants’ eye movements, the study also collected how long the participants

looked at areas within each of the eight mock SERPs (see Appendices U through BB) and

displayed this data as heat maps.

Along with the gaze plots and heat map data, the study also collected video of the

participant’s eye movements and audio from the participants as they viewed the SERPs in the

study. The video clips of individual participants’ eye movements were shown to the participant

during the retrospective think-aloud portions of the study.

Retrospective Think Aloud Data Collection

 28
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

The study also collected qualitative data from the participants using retrospective think-

aloud activities. During this portion of the study, participants were shown video recordings of

their eye-movements as they viewed each mock SERP environment. While they were watching

the videos, they were asked to verbally describe their thought processes as they completed the

task indicated by the user-testing script (see Appendix K). The purpose of collecting this data is

to gain insights into why the participants fixated or had longer fixations on specific locations of

the mock SERP environments as opposed to other areas of the SERP (Bojko, 2013;

Retrospective Think-Aloud | Qualitative Research | Tobii Pro, 2019).

Post-Study Survey

Participants in all groups completed a post-study survey (see Appendix J) after viewing

the mock SERP environments. The post-study survey (see Appendix CC) collected participants’

self-reported preferences of thumbnail designs for each of the eight mock SERP environments.

User Testing Script

The script (see Appendix K) for the eye-tracking study is a modified version of a user

testing script for websites from Steve Krug’s book, “Rocket Surgery Made Easy: The do-it-

yourself guide to finding and fixing usability problems” (Krug, 2010). The prompts for the

retrospective think aloud exercises are modified from the retrospective think-aloud interview

techniques developed by Tobii Pro (Retrospective Think-Aloud | Qualitative Research | Tobii

Pro, 2019).
 29
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Analysis

The eye tracker recorded the participants' eye movements within the mock SERP

environments (see Appendices M through T). This data is represented as gaze plots showing how

each participant's eyes moved through the eight mock SERPs. Next, the gaze plots within each

group were compiled and compared with the compilations of SERP environments across all

groups (see Appendices M through T). These comparisons revealed differences and similarities

in how the participants' eye movements varied between groups and SERP environments. In

addition, this data showed the areas of the mock SERP environments where participants across

all groups looked first. Finally, these comparisons were used to determine if there were varying

gaze patterns between the groups and whether eye movements differed between thumbnail

designs.

The eye tracker also recorded the participants' time spent viewing different areas within

each mock SERP environment (see Appendices U through BB). This data is represented as heat

maps and indicates the areas within the mock SERP environments where the participants spent

the most time viewing. First, the areas within each SERP environment that held the group's gazes

for the longest time were compiled into heat maps for each mock SERP within each group. Next,

these heat maps were compared with the heat maps of the corresponding mock SERP

environments across all three groups (see Appendices U through BB). This comparison revealed

which version of the thumbnail design held the participants' attention the longest in each of the

eight mock SERP environments.

Gaze plots and heat maps were compared with the results of the post-experiment survey

(see Appendix CC). This comparison determined if participants' self-reported preferences for
 30
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

thumbnail designs differed from the design versions that attracted their attention first and held

their attention the longest during the experiment.

The results of the pre-study survey (see Appendix L) established the participants' self-

reported level of knowledge and interest in the topic of each mock SERP environment. In

addition, this information was used to determine if participants' self-reported knowledge and

interest in a specific topic played a role in what area they viewed first within each SERP

environment and how much time they spent viewing the mock SERP.

Finally, the participants' verbal descriptions collected during the retrospective think-aloud

activities were examined to find trends in how participants within each group and among groups

used thumbnails in information-seeking tasks.

 31
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Findings

This study was designed to answer four questions: when users do information-seeking

tasks, do video thumbnail designs with portraits attract users' attention faster than video

thumbnail designs without portraits; when users engage in information-seeking tasks, do video

thumbnail designs with portraits hold users' attention longer than thumbnail designs without

portraits; do users perceive videos with thumbnail designs that include portraits as more

trustworthy than videos with designs that do not have portraits; and will users' show a bias

toward engaging with video content represented by thumbnails with portraits than by thumbnails

without portraits?

This section is divided into four subsections, with each subsection discussing findings

that address the four research questions. The first section addresses if video thumbnails with

portraits attract users' attention faster than designs without portraits. The second section

addresses if thumbnail designs with portraits hold users' attention longer than designs without

portraits. The third section addresses if users perceive thumbnail designs with portraits as more

trustworthy than designs without portraits. And the fourth section address if users show a bias

toward engaging with video content represented by thumbnails with portraits than content

represented by thumbnails without portraits.

Research Question One

The first question this study aims to answer is do video thumbnail designs with portraits

attract users' attention faster than those without portraits. The study collected gaze plots from

each participant, compared participants' eye movements for each SERP between groups, and

found no difference in participants' gaze patterns within or between groups (see Appendices M

through T) If thumbnail designs with portraits attracted participants' attention faster than those
 32
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

without portraits, the gaze plots would have reflected this preference. However, the gaze plots of

the participants in group one, which viewed the target thumbnails in the portrait condition, were

similar to those participants in group two and group three, who viewed the target thumbnails in

the non-portrait and text-only conditions (see Appendices M through T).

However, the gaze plots showed that almost all participants scanned the SERPS in an E

or F-shaped pattern (Pernice, 2017). These findings suggest that the thumbnail design had little

effect on the participants' gaze pattern and that thumbnail designs with portraits do not attract

participants' attention faster than those without portraits. Additionally, the findings suggest that

the position of the thumbnail within the SERP plays a more significant role in attracting

participants' attention first than do the specifics of visuals included in the thumbnail designs.

During the retroactive think-a-loud section of the study, most participants reported that

the titles and descriptions of the videos caught their attention first, followed by the thumbnails.

For example, a participant in group one noted, "I think I just started here, like, I started in the

corner and started scrolling down. Looking at the titles and then pairing with the visuals, but I

did go at the top and went down." gr_001_po_003, personal communication, September 4,

2021). Another participant in group two reported a similar approach when viewing the mock

SERPs, "I first looked at the text of what each video was called. And I looked at the source, that

was the most important thing for me in trying to think through what I wanted to watch was the

title and the source, then I kind of looked back over the images as well" (gr_002_np_005,

personal communication, September 3, 2021). A participant in group three also describes a

similar approach to scanning SERPs, "My thought process was to read all the titles and see

which one grabbed me, and then I look at the picture" (gr_003_to_001, personal communication,

September 1, 2021).
 33
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

One participant had a gaze pattern that stood out from the rest of the participants. This

participant's gaze started in the top left corner of the page but focused on the thumbnail images

first, scanning from the top of the SERP to the bottom, and then shifted to the titles and

descriptions of the videos that piqued their interest. During the retroactive think-a-loud section,

this participant described their approach to scanning the Policy and Higher Education SERP,

"Yeah, so I started looking at the first picture and then reading across to see the title of the video.

And then I didn't feel like reading more, so then; I went off the image alone. And then I jumped

down to the bottom one; I felt like that one looked the most trustworthy, in my opinion. Um, and

then I started reading like the name of that one and then like bouncing over the others to make

sure I was confident in my decision on the last one on the bottom" (gr_003_to_002, personal

communication, August 31). The participant was assigned to group three, saw the target

thumbnail designs in the text-only condition, and scanned all the SERPs similarly.

While the scanning pattern of participant gr_003_to_002 differed slightly from the

scanning pattern of the other participants, in that they scanned thumbnail designs first, rather

than titles and descriptions, the participant still began scanning each SERP in the top left corner

and worked their way down the page. The participant's top-down scanning approach also

suggests that thumbnail designs featuring portraits may not attract participants' attention faster

than thumbnails without portraits. Additionally, it suggests that the position of the thumbnail

design within the SERP may play a more significant role in attracting the participant's attention

first than the design or design elements of the thumbnail.

Research Question Two

The next question this study aims to answer is if video thumbnail designs with portraits

hold users' attention longer than those without portraits. The study collected heatmaps from each
 34
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

participant (see Appendices U through BB) and compared the amount of time participants from

each group held their gaze on the target thumbnail designs within each mock SERP (see

Appendices U through BB). The study’s heatmaps show that thumbnail designs with portraits

held participants' attention longer than thumbnail designs in the non-portrait and text-only

conditions in specific SERPs. However, for other SERPs, non-portrait thumbnail designs held

users' attention for a similar amount of time or slightly longer than thumbnail designs with

portraits or designs in the text-only condition.

For example, heatmaps for the Policy and Higher Education, Too Much Homework,

Black Comedy and Activism, and Sugar Cravings SERPs (see Appendices U; V; X; BB) show

that the thumbnails in the portrait condition held participants' attention longer than the non-

portrait and text-only thumbnail designs. While the heatmap data for the Popular Culture and the

Bible and Randomized Roommates SERPs (see appendices W and Y) show that the thumbnail

designs in the portrait condition and non-portrait condition held the participants' attention for

approximately the same amount of time. The text-only thumbnail designs in both SERPs held the

participants' attention for less time than the portrait and non-portrait thumbnail designs. Finally,

for the Soda Taxes and Is Stress Bad SERPs (see Appendices Z and AA), the non-portrait

thumbnail designs held the participants' attention slightly longer than the portrait thumbnail

designs. Again, the text-only thumbnail designs for both the Soda Tax and Is Stress Bad? SERPs

held the participants' attention for less time than the non-portrait and portrait thumbnail designs.

While heatmap data showed that the amount of time participants spent viewing thumbnails with

portraits compared to non-portrait thumbnails varied from SERP, it cannot account for why these

differences occurred, but it does suggest that SERP topics may play a role in the effectiveness of

thumbnail designs with portraits to hold participants’ attention.

 35
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Additionally, the study’s pre-study survey asked participants to self-report their level of

familiarity with the search topic and their level of interest in learning more about the search topic

(see Appendix L) to gain insight into how participants’ familiarity and insight play a role in what

holds their attention during information search tasks.

The results of the pre-study survey showed that the majority of participants self-reported

being familiar with and interested in learning about the topics associated with the Policy and

Higher Education, Too Much Homework, Black Comedy and Activism, and Sugar Cravings,

SERPs (see Appendix L). However, for the topics associated with Popular Culture and the Bible

and Randomized Roommates SERPs, the majority of participants self-reported being familiar

with topics, but neutral or disinterested in learning more about the topics (see Appendix L).

Compared with the heatmap data from the SERPs, these findings suggest that participants' self-

reported levels of familiarity and interest in learning about the topic may play a role in the

effectiveness of thumbnail designs with portraits versus non-portrait designs in holding their

attention.

Collectively, these findings suggest that both the search topic and the users’ level of

knowledge and interest in learning about the search topic play a role in the effectiveness of

thumbnail designs with portraits holding users’ attention over thumbnail design without portraits

or text-only thumbnail designs.

Research Question Three

The third question this study aims to answer is if users perceive videos with thumbnail

designs that include portraits as more trustworthy than thumbnail designs without portraits. The

study collected data from participants during retroactive think-a-loud sessions to gain insights
 36
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

into if videos represented by thumbnail designs that include portraits are perceived as more

trustworthy than thumbnail designs without portraits.

Most participants considered the trustworthiness of the video's source when selecting to

engage with video content. However, most participants primarily relied on information gleaned

from video titles and descriptions to determine the source's trustworthiness. For example, a

participant in group one describes their thought process for selecting videos. "I'm paying

attention not just to the video image that I see, but also the title, whether there seems to be some

bias in the title, whether the title seems to be pushing me toward a conclusion, and also the

source. I tended to go toward something that seems like a trustworthy, reliable source rather than

just what the image looked like" (gr_001_po_005, personal communication, August 29, 2021).

Another study participant echoes a similar approach of relying more on information from

titles and descriptions to judge the trustworthiness of videos. "I think I was checking out the

source of the videos. But none of the screengrabs really pulled me in at first. So, nothing visually

made sense in this one in particular; I needed to get more information from the text"

(gr_001_po_002, personal communication, August 30, 2021).

While most participants primarily used information gleaned from the title and text

description of the video entries to judge the videos' trustworthiness, some also used information

inferred from the videos' thumbnails to make judgments concerning the reliability of the videos'

sources. For example, a participant in group one, while reviewing the Black Comedy and

Activism SERP (see Appendix F), describes their thought process, "I think it's common that I'm

going to, like, just read all of the headlines. It's going to be what I do first and each time, then

looking at the thumbnails and trying to find something relatable. So, in this case, I was like,

okay, they're talking about race relations, I'm going to not look at the white people… I think, I
 37
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

wanted to see, you know, Black people talk about it." (gr_001_po_001, personal

communications, August 29, 2021). In this example, the participant perceived Black people to be

more trustworthy than non-Black people. Using information gleaned from the videos'

thumbnails, they dismissed videos from sources they perceived as non-Black.

In another example, a study participant, who valued diverse sources and had a desire to

"learn about race-conscious things," eliminated a video because the thumbnail included a portrait

of a person the participant described as "a white man... kind of an old guy" (gr_002_np_001,

personal communication, August 31, 2021). In this example, the participant also used

information from the video's thumbnail design to make judgments about the trustworthiness of

the video's source (see Appendix G).

Most participants cited the trustworthiness and reliability of the source as criteria they

considered their video selection process. However, most participants primarily used video titles

and video descriptions to glean information about the trustworthiness of the sources. However, in

scenarios where participants believed that sources from a particular ethnicity, gender, or age

group, were more trustworthy than other groups, participants showed a bias toward selecting

videos with thumbnails featuring portraits of people belonging to these groups. Additionally,

participants showed a tendency to dismiss videos with thumbnail designs that included portraits

of people not belonging to these groups.

Research Question Four

The final question this study aims to answer is if users show a bias toward engaging with

videos represented by thumbnail designs with portraits than with videos represented by

thumbnail designs without portraits. The results of a post-study survey (see Appendix CC) found

that the participants selected to engage with the thumbnails with portraits for half of the videos in
 38
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

the survey. For the other half of the videos, the participants chose non-portrait thumbnails. The

post-study survey results suggest that participants do not prefer thumbnail designs with portraits

over those without portraits. They also suggest that participants may consider several factors

when selecting video content.

Additionally, during the retroactive think-a-loud sessions, most participants did not show

a bias toward engaging with videos represented by either thumbnail design. Instead, participants

reported primarily using video titles, descriptions, and sources to select a video. If participants

considered thumbnail designs, the designs were used as secondary sources of information and

considered after reading the titles, text, and descriptions.

For example, a participant in group two stated, "I really feel like in general, I'm just

focusing on the words" (gr_002_np_003, personal communication, September 4, 2021) when

describing how they gleaned information about the videos. This same participant continued, "…I

wasn't really paying attention to the thumbnails… I was looking for, I was searching the words

for financial aid" while explaining how they selected a video from the Policy and Higher

Education SERP (gr_002_np_003, personal communication, September 4, 2021).

Another participant in group two described a similar approach, "I'm much more interested

in the title of the video and source" (gr_002_np_005, personal communication, September 4,

2021) when asked if they considered the videos' thumbnail designs when selecting a video.

Finally, a third participant from group one elaborated, "I do think I read the text. I look at the

images too. But I definitely focus more on what the words are saying." (gr_001_po_003,

personal communication, September 4, 2021).

While most participants did not consider thumbnail designs when making selections,

those that did tended to use the designs to gain insights into the style or tone of the video under
 39
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

consideration. For example, a participant in group two inferred from a thumbnail design that a

video under consideration was a documentary-style video (see Appendix G), "I can already tell

from the thumbnail image and the video length that it's a video essay. And I like video essays.

That's like content that I know that I enjoy. So, I'm not taking a boredom risk. Whereas I'd be

definitely taking a boredom risk with the other four." (gr_002_np_004, personal communication,

August 31, 2021).

Another participant in group two describes similarly using thumbnails while selecting a

video from the Sugar Cravings SERP (see Appendix G), "The visuals really stuck out to me

because they feel like some of the strongest indicators (of how this topic could be approached).

Like, you could take this topic in so many different directions that the visuals really were

required to help my brain organize it. I did a quick scan and was like I don't want talk show

vibes, so I don't want four and five because I don't just want general chatter about this."

(gr_002_np_002, personal communication, September 3, 2021. In this case, the participant

inferred from the thumbnail designs of the videos in positions four and five were from a

television talk show. Unlike the previous example, this participant eliminated those videos as

options due to their dislike of television talk show videos.

A third example occurred when a participant dismissed videos because they inferred from

the thumbnail designs that the videos were produced in a television news program style (see

appendix QQ). "I tend to try and avoid news channels. I feel like a lot of news channels are

fearmongering or they do a lot of talk pieces that aren't very relevant. So, when I'm looking at

YouTube clips, I don't really go for news channels…I also liked the thumbnail for the Duke

University one, and that was also not a talk show." (gr_002_np_001, personal communication,

August 31, 2021).

 40
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Most participants primarily relied on videos' titles, descriptions, and sources when

selecting videos. However, some participants relied on video thumbnails as a secondary source

of information about the videos under consideration. Participants did not appear to be biased

toward using thumbnail designs with portraits over those without portraits. However, most

participants used information from thumbnail designs to determine the style or tone of the video

under consideration.
 41
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Implications

The findings of the study suggest that thumbnail designs with portraits do not attract

users' attention faster than designs without portraits. However, in certain information-seeking

tasks, thumbnail designs with portraits hold users' attention longer than designs without portraits.

In other information-seeking tasks, designs without portraits hold users' attention longer than

designs with portraits. These findings suggest that search topics, users' level of knowledge, and

interest in the search topic may play a role in the effectiveness of thumbnail designs with

portraits to attract user’s attention faster than non-portrait designs.

Additionally, users tend to use thumbnail designs as a secondary source of information

when assessing the trustworthiness of video content. However, when users value information

from a source belonging to a particular group or identity, users perceive designs with portraits of

people belonging to the desired group or identity as more trustworthy than designs without

portraits. Finally, users did not show a bias toward engaging with videos represented by

thumbnail designs with portraits over designs without portraits. Instead, users reported using

video titles, descriptions, and sources as primary sources of information when selecting content.

Thumbnail designs were used as secondary sources of information and typically to determine the

style or tone of the video content.

These findings suggest that designers should consider how potential users may interact

with the designs when viewed within the context of SERP environments. Designers often have

little influence on the exact placement of the thumbnail design within a SERP environment.

However, they should consider the possible search scenarios in which users may interact with the

design and potential users' level of knowledge about the topic when selecting visuals to feature in

the thumbnail.
 42
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Limitations

The researcher acknowledges the limitations of this study. The study was conducted to

provide insight into an area that currently has very little academic research. The study's small

participant sample consisted of a largely homogeneous population (participants residing in the

Research Triangle region of North Carolina). The study did not collect information about the

participant's expertise or familiarity with digital information-seeking tasks, and users' experience

levels could affect how they approach search activities. Additionally, the study did not account

for specific age ranges of the participants. It is possible that users who identify as digital natives

might differ from those who identify as non-digital natives or digital immigrants.

The mock SERP environments were designed to mimic a desktop computer screen. SERP

environments designed for mobile devices might have produced different results. Also, the target

thumbnail designs (see Appendices C through E) were similar in design. Some participants

mentioned that they recognized the similarly designed thumbnails that appeared throughout the

test. Therefore, participants' behavior might have been affected by identifying the similarity of

the target thumbnail designs.

Additionally, the testing took place in various locations, including outside environments,

where naturally occurring changes in the ambient light levels may have contributed to errors in

the eye-tracker's ability to collect and record data. For example, during the testing phase of the

study, the eye-tracking equipment did not record data for the third participant of the first group

for the Sugar Cravings SERP. The eye-tracker also did not record data for the first, second,

fourth, and fifth participants in the third group for the Randomized Roommate SERP. The reason

for these errors in data collection is unknown; however, the fluctuations in lighting at the testing

sites might have been a contributing cause.

 43
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

The findings of this study are specific to this study and cannot be extrapolated or

generalized. However, the results are intended to provide a foundation for further research into

how users engage with visuals, specifically video thumbnails, when selecting content during

information-seeking tasks. Additionally, the findings are intended to be used as a foundation for

creating best practices guidelines for content creators and social media managers when designing

video thumbnails that will be used to aid in information-seeking tasks.

 44
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Conclusion

The study investigated if thumbnail designs with portraits are more effective in drawing

users’ attention and holding it during information search tasks than thumbnail designs without

portraits. While designs with portraits did not draw users’ attention faster than designs without

portraits, designs with portraits held users’ attention longer than designs without portraits during

some information-seeking tasks.

Additionally, the study also investigated if users perceive designs with portraits as more

trustworthy than those without portraits. The findings suggest that users primarily relied on video

titles, descriptions, and sources to assess the trustworthiness of the content. However, study’s

findings suggest that in some search scenarios, users believed that information from sources

belonging to a particular ethnicity, gender, age, or identity, was more trustworthy than

information from other sources. In these scenarios, users showed a biased in selecting thumbnail

designs with portraits of individuals they perceived as belonging to the more trustworthy identity

or group.

The study findings suggest that users are not biased toward engaging with content

represented by thumbnail designs with portraits over designs without portraits. Additionally, the

findings suggest that users primarily rely on video titles, descriptions, and sources when

selecting content and use thumbnail designs as a secondary sources of information to determine

the style or tone of the video content.

 45
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Further Research

Currently, there is very little literature exploring how or if factors, such as visuals,

influence users' viewing behavior on SERPs. This study is meant to serve as a pilot study for

further research into how thumbnail design, specifically those that incorporate portraits, may

influence users' viewing behavior during information-seeking tasks.

Further research should examine how SERP designs, such as lists, grids, or other alternative

layouts, may affect users' scanning patterns. Additionally, research is needed into how the device

type, desktop, tablet, or mobile, may play a role in influencing users' viewing behavior of

SERPs.

Additional research should examine how the users' familiarity with search topics might

affect their viewing behavior and should attempt to control for these factors by using a

standardized set of tasks. Future research should also consider how individual characteristics

such as users' age, education level, and prior knowledge about the search topic may affect their

viewing behavior of SERPs. Finally, future studies should also consider how individual users'

attitudes and feelings toward search topics may affect how they view SERPs.
 46
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix A

Link to high-res files: https://duke.box.com/s/mabuipwh03qmdibgmy8ocu3bzvhw0z1e

Mock SERP for higher education policy search Mock SERP for too much homework search

Mock SERP for pop culture and Bible search Mock SERP for Black comedy and activism search

Mock SERP for randomized roommates’ search Mock SERP for soda tax search

Mock SERP Is stress bad? search Mock SERP sugar cravings search
 47
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix B

Video titles from the “Extra Credit” video series were used as search terms to create mock SERP

environments.

Extra Credit: Life’s Curiosities Explained Search Terms used to create mock SERPs
Video Titles
Policy & Higher Education higher education policy
How Much Homework Is Too Much? too much homework
Pop Culture & The Bible pop culture and Bible
Black Comedy & Activism Black comedy and activism
Randomly Assigned Roommates: What Are randomized roommates
The Effects?
Soda Taxes? soda tax
Is Stress Always Bad? Is stress bad?
Why Do I Crave Sugar? sugar cravings
 48
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix C

Group One Thumbnail Designs with Portraits

Link to high-res files: https://duke.box.com/s/tin3630x0nxj7kea08dygl0xne1gebyl

Policy and higher education video thumbnail design Too much homework video thumbnail design featuring
featuring a portrait of the expert (Wolf, n.d.-e) a portrait of the expert (Wolf, n.d.-b)

Pop culture and the Bible video thumbnail design Black comedy and activism video thumbnail design
featuring a portrait of the expert (Wolf, n.d.-d) featuring a portrait of the expert (Wolf, n.d.-a)

Randomly assigned roommate video thumbnail design Soda taxes video thumbnail design featuring a portrait of the
featuring a portrait of the expert (Wolf, n.d.-f) expert (Wolf, n.d.-g)

Is stress bad? thumbnail design featuring a portrait of the Why do I crave sugar? thumbnail design featuring a portrait
expert (Wolf, n.d.-c) of the expert (Wolf, n.d.-h)
 49
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix D

Group Two Thumbnail Designs with Non-Portraits

Link to high-res files: https://duke.box.com/s/smmqf42dquixjaat0qfyrcxmbooziepz

Policy and higher education thumbnail design featuring a Too much homework thumbnail design featuring a non-
non-portrait visual (Academy, Book, Building, College, portrait visual (Homework - Homework Icon Black And White
School, Structure - World Class Facilities Icon - Full Size - Full Size PNG Clipart Images Download. ClipartMax.Com,
PNG Clipart Images Download. ClipartMax.Com, n.d.; n.d.; Wolf, n.d.-b)
Wolf, n.d.-h)

Pop culture and the Bible thumbnail design featuring a non- Black comedy and activism thumbnail design featuring a
portrait visual (Clipart - Bible Clipart - Full Size PNG non-portrait visual (Black Panther Clipart Black Power -
Clipart Images Download. ClipartMax.Com, n.d.; Wolf, n.d.- Raised Fist Clip Art - Free Transparent PNG Clipart Images
d) Download, n.d.; Wolf, n.d.-a)

Randomly assigned roommate thumbnail design featuring a Soda taxes thumbnail design featuring a non-portrait visual
non-portrait visual (Need A Roommate - Illustration - Free (Soda Bottle Vector - Soda Bottle Icon - Free Transparent
Transparent PNG Clipart Images Download, n.d.; Wolf, PNG Clipart Images Download, n.d.; Wolf, n.d.-g)
n.d.-f)

Is stress bad? thumbnail design featuring a non-portrait visual (Stress Why do I crave sugar? thumbnail design featuring a non-portrait
Comments - Stress Symbol - Full Size PNG Clipart Images visual (Sugar Clipart Png - Bowl Of Sugar Clipart - Full Size PNG
Download. ClipartMax.Com, n.d.; Wolf, n.d.-c) Clipart Images Download. ClipartMax.Com, n.d.; Wolf, n.d.-h)
 50
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix E

Group Three Thumbnail Designs with Text Only

Link to high-res files: https://duke.box.com/s/z4k3hgp8nysdmy54eqrvh2oz3shcp2lo

Policy and higher education text only thumbnail Too much homework text only thumbnail design
design (Wolf, n.d.-e). (Wolf, n.d.-b)

Pop culture and the Bible text only thumbnail Black comedy and activism text only thumbnail
design (Wolf, n.d.-d). design (Wolf, n.d.-a).

Randomly assigned roommates text only Soda taxes text only thumbnail design (Wolf, n.d.-
thumbnail design (Wolf, n.d.-f). g).

Is stress bad? text only thumbnail design (Wolf, Why do I crave sugar? text only thumbnail design
n.d.-c). (Wolf, n.d.-h).
 51
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix F

Mock SERPs with Thumbnail Designs with Portraits

Link to high-res files: https://duke.box.com/s/unpj5w067rqr4vc2687ay0qkmdjy0dfx

Participant group one; policy and higher Participant group one; too much homework
education mock SERP; portrait thumbnail mock SERP; portrait thumbnail

Participant group one; pop culture and the Participant group one; Black comedy and
Bible mock SERP; portrait thumbnail activism mock SERP; portrait thumbnail

Participant group one; randomized roommates Participant group one; soda tax mock SERP;
mock SERP; portrait thumbnail portrait thumbnail

Participant group one; is stress bad? mock Participant group one; why do I crave sugar?
SERP; portrait thumbnail mock SERP; portrait thumbnail
 52
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix G

Mock SERPs with Thumbnail Designs with Non-Portraits

Link to high-res files: https://duke.box.com/s/jjn4f0a7chlzunkthjd91umtfoh1albu

Participant group two; policy and higher Participant group two; too much homework mock
education mock SERP; non-portrait thumbnail SERP; non-portrait thumbnail

Participant group two; pop culture and the Bible Participant group two; Black comedy and
mock SERP; non-portrait thumbnail activism mock SERP; non-portrait thumbnail

Participant group two; randomized roommates Participant group two; soda tax mock SERP; non-
mock SERP; non-portrait thumbnail portrait thumbnail

Participant group two; is stress bad? mock SERP; Participant group two; why do I crave sugar mock
non-portrait thumbnail SERP; non-portrait thumbnail
 53
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix H

Mock SERPs with Thumbnail Designs with Text-Only

Link to high-res files: https://duke.box.com/s/l39ugrgn9bqiec7j4kurzds6ivc1verc

Participant group three; policy and higher Participant group three; too much homework
education mock SERP; text only thumbnail mock SERP; text only thumbnail

Participant group three; pop culture and the Bible Participant group three; Black comedy and
mock SERP; text only thumbnail activism mock SERP; text only thumbnail

Participant group three; randomized roommates Participant group three; soda tax mock SERP; text
mock SERP; text only thumbnail only thumbnail

Participant group three; is stress bad? mock Participant group three; why do I crave sugar
SERP; text only thumbnail mock SERP; text only thumbnail
 54
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix I

Pre-Study Survey

https://unc.az1.qualtrics.com/jfe/form/SV_eb8mYPdnimTMIFU
 55
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix J

Post-Study Survey

https://unc.az1.qualtrics.com/jfe/form/SV_4UvROxwCQjBg8XY
 56
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix K

Script for Eye-tracking data collection experiments for group one, two and three. The script is a

modified version of a user testing script for websites from Steve Krug’s book, “Rocket Surgery

Made Easy: The do-it-yourself guide to finding and fixing usability problems” (Krug, 2010).

Eye-Tracking Experiment Script

The Instructions

“Hi, my name is Megan Mendenhall, and I’m a graduate student at the Hussman School
of Journalism and Media at UNC. Thank you for your interest in this study.
Before we begin, I have some information for you, and I’m going to read it to make sure
that I cover everything. I’m asking people to view a series of video thumbnail designs as they
appear within search engine result pages so I can get information on how to optimize designs for
video thumbnails. The session should take about 20 to 40 minutes.
The first thing I want to make clear right away is that I’m testing the design of the video
thumbnails, not you. You can’t do anything wrong here. In fact, this is probably the one place
today where you don’t have to worry about making mistakes.
Also, please don’t worry that you’re going to hurt my feelings. I’m doing this to improve
the designs of video thumbnails, so I need your honest reactions.
If you have any questions as we go along, just ask them. I may not be able to answer
them right away, since I’m interested in how people react to the designs when they don’t have
someone sitting next to them to help. If you still have any questions when we’re done, I’ll try to
answer them then. And if you need to take a break at any point, just let me know.
You may have noticed the eye-tracking equipment. With your permission, I’m going to
record what and where your eyes look as you review eight search engine results pages. The data
will only be used in this research project, and it won’t be seen by anyone except the people
working on this project.
If you are interested in participating in the study, please let me know at this time and I
will provide a consent form for you to review. If you are no longer interested in participating in
the study, please let me know at this time.”

[If a potential participant indicates that they are no longer interested in participating in the study,
the researcher will thank them for their time and interest and the interaction will end at this point.
If a potential subject indicates that they are interested in participating in the study, the researcher
will provide them with a paper copy of the consent form.]

[Researcher provides potential participant with a paper copy of the consent form.]
 57
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

“I’m going to read the consent form out loud to you, but in summary, it says that I have
your permission to record your eye movements, and that the recordings will only be seen by the
people working on the project.”

[The researcher reads the consent form out load to all potential subjects.]

“Do you have any questions or concerns?”

[Researcher answers questions and addresses concerns from potential participants]

“Please take your time to review the consent form. If you are interested in participating in
the study, please sign the consent form and hand it back to me.”

[If a potential participant indicates that they are no longer interested in participating in the study,
the researcher will thank them for their time and interest and the interaction will end at this point.
If a potential subject indicates that they are interested in participating in the study, the researcher
will ask them to sign the consent form at this point.]

“Please sign the consent form.”

[Researcher collects the signed consent form and provides the participant with a signed copy of
their consent form]

“OK. Before we begin the experiment, I’d like to ask you to take a short pre-experiment
survey. After you complete the pre-experiment survey you will see eight search engine result
pages.”

[Participants in all groups take the pre-experiment survey questions (see Appendix FFF).
The answers will be recorded digitally.]

[Researcher opens the window to the digital Pre-Experiment Survey]

“Please complete the pre-survey. When you have completed it, let me know.”

[Participants completes the Pre-Experiment Survey]

“Thanks, we will move to the next section of the experiment where you will view search
engine result pages. Before we start, we’ll need to calibrate the hardware.”

[Researcher opens eye tracking software and positions the participant so that the eye-tracking
equipment can calibrate the participant’s eyes]
 58
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

“During the calibration process, you will see red calibration point appear on the screen.
Please follow the calibration points with your eyes as they moves across the screen. The
calibration will automatically stop when all calibration points have been shown.”

[researcher begins the calibration]

[Researcher verifies the successful calibration of the eye tracker]

“Thanks, the eye tracker is now calibrated.”

[Researcher opens the software to full-screen mode and begins the test.]

“In this section of the experiment, you will view eight search engine result pages. When
you are ready, you may press the space bar to advance to the next screen.”

“OK, imagine you are seeking information about state and federal financial aid programs
designed to assist college students with educational expenses. Look over the results on the next
screen and make note of the video that you are most likely to select. When you are ready, you
may press the space bar to advance to the next screen.”

The search engine returns the following results:

[Group one participants view appendix HH]
[Group two participants view appendix PP]
[Group three participants view appendix XX]

[Participant’s eye-tracking data is collected for appendix HH or PP or XX]

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“Ok, imagine you have a second-grade niece who is assigned 40 minutes of homework
every school night. You are seeking information about the appropriate amount of homework for
second-grade students. Look over the results on the next screen and make note of the video that
you are most likely to select. When you are ready, you may press the space bar to advance to the
next screen.”

The search engine returns the following results:

[Group one participants view appendix II]
[Group two participants view appendix QQ]
[Group three participants view appendix YY]

[Participant’s eye-tracking data is collected for appendix II or QQ or YY]

 59
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you just finished watching the classic dark-comedy crime film Pulp
Fiction. Throughout the movie, Jules Winnfield, one of the main characters, recites what he
describes as a biblical passage, Ezekiel 25:17. You are interested in learning if other popular
movies, music, and books also reference the Bible. Look over the results on the next screen and
make note of the video that you are most likely to select. When you are ready, you may press the
space bar to advance to the next screen.”

The search engine returns the following results:

[Group one participants view appendix JJ]
[Group two participants view appendix RR]
[Group three participants view appendix ZZ]

[Participant’s eye-tracking data is collected for appendix JJ or RR or ZZ]

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you saw an episode of "United Shades of America," a documentary

television series hosted by comedian and political activist W. Kamau Bell that tackles racism and
immigration. You are interested in learning about other entertainers who use comedy to explore
human and civil rights issues. Look over the results on the next screen and make note of the
video that you are most likely to select. When you are ready, you may press the space bar to
advance to the next screen.”

The search engine returns the following results:

[Group one participants view appendix KK]
[Group two participants view appendix SS]
[Group three participants view appendix AAA]

[Participant’s eye-tracking data is collected for appendix KK or SS or AAA]

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you are going to be a first-year student at a university next fall. You are
filling out a housing application for the upcoming year and you have the choice of being
 60
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

assigned a roommate or choosing your own roommate. You are interested in learning more about
the pros and cons of living with an assigned roommate versus a friend. Look over the results on
the next screen and make note of the video that you are most likely to select. When you are
ready, you may press the space bar to advance to the next screen.”

The search engine results return the following results.

[Group one participants view appendix LL]
[Group two participants view appendix TT]
[Group three participants view appendix BBB]

[Participant’s eye-tracking data is collected for appendix LL or TT or BBB]

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you are voting in an upcoming county election. One of the initiatives on
the ballot is a tax on soft drinks and other sweetened beverages. Other cities passed similar
beverage taxes, and you are interested in learning about the outcomes of these taxes. Look over
the results on the next screen and make note of the video that you are most likely to select. When
you are ready, you may press the space bar to advance to the next screen.”

The search engine returns the following results:

[Group one participants view appendix MM]
[Group two participants view appendix UU]
[Group three participants view appendix CCC]

[Participant’s eye-tracking data is collected for appendix MM or UU or CCC]

“Once you have made your selection, you may press the space bar to advance to the next
screen.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you are a high-school student taking four advanced placement courses at
the same time. You want to manage the stress of a heavy workload by learning about the positive
and negative effects of stress. Look over the results on the next screen and make note of the
video that you are most likely to select. When you are ready, you may press the space bar to
advance to the next screen.”

The search engine returns the following results:

[Group one participants view appendix NN]
[Group two participants view appendix VV]
[Group three participants view appendix DDD]
 61
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

[Participant’s eye-tracking data is collected for appendix NN or VV or DDD]

“Once you have made your selection, you may press the space bar to advance to the next
page.”

[Participant presses the space bar to advance to the next screen.]

“OK, imagine you are training for a 5K race, and as part of your training routine, you
want to limit the amount of junk food in your diet. You have a sweet tooth and curbing your
sugar habit will be difficult. You want to understand these cravings better so that you can have
control of your diet. Look over the results on the next screen and make note of the video that you
are most likely to select. When you are ready, you may press the space bar to advance to the next
screen.”

The search engine returns the following results:

[Group one participants view appendix OO]
[Group two participants view appendix WW]
[Group three participants view appendix EEE]

[Participant’s eye-tracking data is collected for appendix OO or WW or EEE]

“Once you have made your selection, you may press the space bar to advance to the next
page.”

[Participant presses the space bar to advance to the next screen.]

“Thank you, you have completed this portion of the experiment.”

[Researcher saves the recording]

“Now I’m going to play videos of your eye movements as you looked the eight search
engine result pages. As you watch each recording, you will see a red points and lines appear over
the search engine result pages. The points represent where your eyes stopped to look, and the red
lines represent the direction of you gaze. As we review each video, describe your thoughts as you
looked through the search engine result page and made your selection.”

[Researcher cues the computer to record a Retro Active Think-A-Loud video]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

HH/PP/XX]

“In this video, you viewed a search engine results page for information about state and
federal financial aid programs designed to assist college students with educational expenses. As
 62
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

you watch the video, describe your thoughts as you looked through the page and made your
selection.”

[Participants in group one will watch their eye-movements while viewing appendix HH]
[Participants in group two will watch their eye-movements while viewing appendix PP]
[Participants in group three will watch their eye-movements while viewing appendix XX]

[Participant describes thought process as he or she viewed appendix HH/PP/XX]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

II/QQ/YY]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about the appropriate amount of homework for second-
grade students. As you watch the video, describe your thoughts as you looked through the page
and made your selection.”

[Participants in group one will watch their eye-movements while viewing appendix II]
[Participants in group two will watch their eye-movements while viewing appendix QQ]
[Participants in group three will watch their eye-movements while viewing appendix YY]

[Participant describes thought process as he or she viewed appendix II/QQ/YY]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

JJ/RR/ZZ]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about popular movies, music, and books that reference the
Bible. As you watch the video, describe your thoughts as you looked through the page and made
your selection.”
 63
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

[Participants in group one will watch their eye-movements while viewing appendix JJ]
[Participants in group two will watch their eye-movements while viewing appendix RR]
[Participants in group three will watch their eye-movements while viewing appendix ZZ]

[Participant describes thought process as he or she viewed appendix JJ/RR/ZZ]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

KK/SS/AAA]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about entertainers who use comedy to explore human and
civil rights issues. As you watch the video, describe your thoughts as you looked through the
page and made your selection.”

[Participants in group one will watch their eye-movements while viewing appendix KK]
[Participants in group two will watch their eye-movements while viewing appendix SS]
[Participants in group three will watch their eye-movements while viewing appendix AAA]

[Participant describes thought process as he or she viewed appendix KK/SS/AAA]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

LL/TT/BBB]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about the pros and cons of living with an assigned
roommate versus a friend. As you watch the video, describe your thoughts as you looked through
the page and made your selection.”

[Participants in group one will watch their eye-movements while viewing appendix LL]
[Participants in group two will watch their eye-movements while viewing appendix TT]
[Participants in group three will watch their eye-movements while viewing appendix BBB]
 64
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

[Participant describes thought process as he or she viewed appendix LL/TT/BBB]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

MM/UU/CCC]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about the outcomes of beverage taxes. As you watch the
video, describe your thoughts as you looked through the page and made your selection.”

[Participants in group one will watch their eye-movements while viewing appendix MM]
[Participants in group two will watch their eye-movements while viewing appendix UU]
[Participants in group three will watch their eye-movements while viewing appendix CCC]

[Participant describes thought process as he or she viewed appendix MM/UU/CCC]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

NN/VV/DDD]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about the positive and negative effects of stress. As you
watch the video, describe your thoughts as you looked through the page and made your
selection.”

[Participants in group one will watch their eye-movements while viewing appendix NN]
[Participants in group two will watch their eye-movements while viewing appendix VV]
[Participants in group three will watch their eye-movements while viewing appendix DDD]

[Participant describes thought process as he or she viewed appendix NN/VV/DDD]

[Researcher may ask participant to elaborate]

 65
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”
[Participant’s responses are recorded]

[Researcher sets program to display recording of participant viewing thumbnails in appendix

OO/WW/EEE]

“Great, now you will see a recording of your eye movements as you viewed the search
engine results page with information about sugar cravings. As you watch the video, describe
your thoughts as you looked through the page and made your selection.”

[Participants in group one will watch their eye-movements while viewing appendix OO]
[Participants in group two will watch their eye-movements while viewing appendix WW]
[Participants in group three will watch their eye-movements while viewing appendix EEE]

[Participant describes thought process as he or she viewed appendix OO/WW/EEE]

[Researcher may ask participant to elaborate]

“Which thumbnail caught your eye first and why? Which thumbnail did you select and
why?”

[Participant’s responses are recorded]

“Great, thank you, we have completed this section of the experiment. I will save the
recording.”

[Researcher saves the recording]

“Now I’m going to ask you to complete a post experiment survey about the search engine
results pages you have just seen. When you have completed the survey, we will be finished with
the experiment. Let me know when you have completed the survey.”

[Researcher opens the window to the digital Post Experiment Survey]

[Participants in all groups take the post experiment survey (see Appendix GGG). The answers
are recorded digitally.]

[Participant indicates that he or she has completed the survey.]

Wrapping Up

“Do you have any questions for me now that we’re done? Thank you for your time and
participation in this project.”
 66
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix L

Pre-study survey results for Higher Education and Policy statements

Statement one: I am familiar with financial aid programs to assist college students with
educational expenses.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 5 of 15
Somewhat agree: 8 of 15
Neither agree nor disagree: 0 of 15
Somewhat disagree: 2 of 15
Strongly disagree: 0 of 15
 67
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about financial aid programs to assist college students
with educational expenses.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 6 of 15
Somewhat agree: 3 of 15
Neither agree nor disagree: 3 of 15
Somewhat disagree: 2 of 15
Strongly disagree: 1 of 15

• Majority of participants (13 of 15) self-reported familiarity with topic.

• Nine participants self-reported interest in topic, 6 participants reported no interest nor
disinterest with the topic.
 68
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Too Much Homework statements

Statement one: I am familiar with the issues surrounding the appropriate amount of homework
for elementary school students.

Participants Strongly agree Somewhat Neither agree Somewhat Strongly

agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 2 of 15
Somewhat agree: 8 of 15
Neither agree nor disagree: 1 of 15
Somewhat disagree: 2 of 15
Strongly disagree: 2 of 15
 69
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about the issues surrounding the appropriate amount
of homework for elementary school students.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 3 of 15
Somewhat agree: 8 of 15
Neither agree nor disagree: 2 of 15
Somewhat disagree: 2 of 15
Strongly disagree: 0 of 15

• Majority of participants (10 of 15) self-reported being familiar with the topic
• Majority of participants (11 of 15) self-reported being interested in learning more about
the topic.
 70
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Popular Culture and the Bible statements

Statement one: I am familiar with biblical references in popular movies, music, and books.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 3 of 15
Somewhat agree: 7 of 15
Neither agree nor disagree: 1 of 15
Somewhat disagree: 4 of 15
Strongly disagree: 0 of 15
 71
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about biblical references in popular movies, music,
and books.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x x
Part. Two
Group Two,
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 1 of 15
Somewhat agree: 5 of 15
Neither agree nor disagree: 5 of 15
Somewhat disagree: 3 of 15
Strongly disagree: 1 of 15

• Majority of participants (10 of 15) self-reported familiarity with the topic.

• Majority of participants (9 of 15) self-reported neutral or disinterest in learning about the
topic.
 72
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Black comedy and activism statements

Statement one: I am familiar with entertainers who use comedy to explore human and civil rights
issues.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 3 of 15
Somewhat agree: 11 of 15
Neither agree nor disagree: 0 of 15
Somewhat disagree: 1 of 15
Strongly disagree: 0 of 15
 73
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about entertainers who use comedy to explore human
and civil rights issues.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 4 of 15
Somewhat agree: 8 of 15
Neither agree nor disagree: 2 of 15
Somewhat disagree: 1 of 15
Strongly disagree: 0 of 15

• Majority of participants (14 of 15) self-reported familiarity with the topic.

• Majority of participants (12 of 15) self-reported interest in learning more about the topic.
 74
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Randomized Roommates statements

Statement one: I am familiar with issues surrounding living with an assigned roommate versus a
friend.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 2 of 15
Somewhat agree: 6 of 15
Neither agree nor disagree: 3 of 15
Somewhat disagree: 3 of 15
Strongly disagree: 1 of 15
 75
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about the issues living with an assigned roommate
versus a friend.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 1 of 15
Somewhat agree: 2 of 15
Neither agree nor disagree: 6 of 15
Somewhat disagree: 4 of 15
Strongly disagree: 2 of 15

• Over half of the participants (8 of 15) self-reported familiarity with topic.

• Majority of participants (12 of 15) self-reported neutral or disinterest in learning about the topic.
• Pre experiment survey results for Soda Tax statements
 76
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement one: I am familiar with soda taxes.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 0 of 15
Somewhat agree: 6 of 15
Neither agree nor disagree: 1 of 15
Somewhat disagree: 7 of 15
Strongly disagree: 1 of 15
 77
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about soda taxes.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 1 of 15
Somewhat agree: 4 of 15
Neither agree nor disagree: 2 of 15
Somewhat disagree: 4 of 15
Strongly disagree: 4 of 15

• Majority (8 of 15) neutral knowledge levels or unfamiliarity about the topic.

• Majority (10 of 15) report neutral or disinterest in learning about the topic.
 78
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Is Stress Bad statements

Statement one: I am familiar with the effects of stress.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 10 of 15
Somewhat agree: 5 of 15
Neither agree nor disagree: 0 of 15
Somewhat disagree: 0 of 15
Strongly disagree: 0 of 15
 79
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about the effects of stress.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 5 of 15
Somewhat agree: 10 of 15
Neither agree nor disagree: 0 of 15
Somewhat disagree: 0 of 15
Strongly disagree: 0 of 15

• All participants (15 of 15) self-reported familiarity with the topic.

• All participants (15 of 15) self-report interest in learning about the topic.
 80
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Pre-study survey results for Sugar Cravings statements

Statement one: I am familiar with the challenges of limiting the amount of junk food in your diet.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

Group One, x
Part. One
Group One, x
Part. Two
Group One, x
Part. Three
Group One, x
Part. Four
Group One, x
Part. Five

Group Two, x
Part. One
Group Two, x
Part. Two
Group Two, x
Part. Three
Group Two, x
Part. Four
Group Two, x
Part. Five

Group Three, x
Part. One
Group Three, x
Part. Two
Group Three, x
Part. Three
Group Three, x
Part. Four
Group Three, x
Part. Five

Strongly agree: 7 of 15
Somewhat agree: 7 of 15
Neither agree nor disagree: 1 of 15
Somewhat disagree: 0 of 15
Strongly disagree: 0 of 15
 81
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Statement two: I am interested in learning about the challenges of limiting the amount of junk
food in your diet.

Participants Strongly Somewhat Neither agree Somewhat Strongly

agree agree nor disagree disagree disagree

gr_001_po_001 x

gr_001_po_002 x

gr_001_po_003 x

gr_001_po_004 x

gr_001_po_005 x

gr_002_np_001 x

gr_002_np_002 x

gr_002_np_003 x

gr_002_np_004 x

gr_002_np_005 x

gr_003_to_001 x

gr_003_to_002 x

gr_003_to_003 x

gr_003_to_004 x

gr_003_to_005 x

Strongly agree: 3 of 15
Somewhat agree: 7 of 15
Neither agree nor disagree: 3 of 15
Somewhat disagree: 1 of 15
Strongly disagree: 1 of 15

• Majority of participants (14 of 15) self-reported familiarity with the topic.

• Majority of participants (10 of 15) self-reported interest in learning more about the topic.
 82
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix M

Policy and Higher Education mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/a3mcPgn2TGg
 83
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix N

Too Much Homework mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/9aUalPwsOP0
 84
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix O

Popular culture and Bible mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/BPEsp0sJLPA
 85
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix P

Black comedy and activism mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/cNfUKWeLgMU
 86
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix Q

Randomized Roommates mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/EL0qD9BGlCo
 87
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix R

Soda Taxes mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/Rd_X6pDVID0
 88
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix S

Is Stress Bad? Mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/ZTKTYYWgx3s
 89
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix T

Sugar Cravings mock SERP Gaze Plots of Group One, Two and Three participants

https://youtu.be/ZeR-iwm65EA
 90
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix U

Policy and Higher Education Policy mock SERP Heat Maps of Group One, Two and Three

participants

Link to high-res files: https://duke.box.com/s/en9lk61fuflahjpdd2nvppm51il5m64r

 91
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix V

Too Much Homework mock SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/t5gltyq8zv95kss6fovr8ff6bqk84tjz

 92
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix W

Popular Culture and Bible SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/mtyiweunosg0m21rr1ifa2l9082b5ggl

 93
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix X

Black Comedy and Activism SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/wmeiq6kk2ukbnlakp1ik103q7ij56gc1

 94
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix Y

Randomized Roommates SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/sxtew2ydz5zm8mkivf4wpkax2pbwj4nr

 95
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix Z

Soda Taxes SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/w80gotom1dz5y9nvld1933yfmoklw122

 96
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix AA

Is Stress Bad? SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/xp6k70mlnwtsd2uq2j0svznf4mymbcr7

 97
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix BB

Sugar Cravings SERP Heat Maps of Group One, Two and Three participants

Link to high-res files: https://duke.box.com/s/27bl698bzd7zduu5usoty48mih3u5x8n

 98
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Appendix CC

Post-study survey results for Higher Education and Policy thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 10 of 15
Non-portrait design: 5 of 15
Text-only design: 0 of 15
 99
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Too Much Homework thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 8 of 15
Non-portrait design: 7 of 15
Text-only design: 0 of 15
 100
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Pop culture and Bible thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 5 of 15
Non-portrait design: 8 of 15
Text-only design: 2 of 15
 101
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Black comedy and activism thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 8 of 15
Non-portrait design: 6 of 15
Text-only design: 1 of 15
 102
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Randomized Roommates thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 1 of 15
Non-portrait design: 14 of 15
Text-only design: 0 of 15
 103
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Soda tax thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 3 of 15
Non-portrait design: 12 of 15
Text-only design: 0 of 15
 104
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Is Stress Bad thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 9 of 15
Non-portrait design: 6 of 15
Text-only design: 0 of 15
 105
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Post-study survey results for Sugar Cravings thumbnail designs

Participants Portrait design Non-portrait design Text-only design

thumbnail thumbnail thumbnail

Group One, Part. One x

Group One, Part. Two x

Group One, Part. Three x

Group One, Part. Four x

Group One, Part. Five x

Group Two, Part. One x

Group Two, Part. Two x

Group Two, Part. Three x

Group Two, Part. Four x

Group Two, Part. Five x

Group Three, Part. One x

Group Three, Part. Two x

Group Three, Part. Three x

Group Three, Part. Four x

Group Three, Part. Five x

Portrait design: 2 of 15
Non-portrait design: 10 of 15
Text-only design: 3 of 15
 106
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

References

Academy, Book, Building, College, School, Structure—World Class Facilities Icon—Full Size

PNG Clipart Images Download. ClipartMax.com. (n.d.). Retrieved September 28, 2020,

from https://www.clipartmax.com/middle/m2i8H7Z5b1K9N4G6_academy-book-

building-college-school-structure-world-class-facilities-icon/

Agbbon, A. (2019, February 27). Video Thumbnails and Their Significance in Video Marketing |

WowMakers. WowMakers Blog | Thoughts on User Experience Design, Video

Marketing, Branding and More. https://www.wowmakers.com/blog/video-thumbnails-in-

video-marketing/

Aharon, I., Etcoff, N., Ariely, D., Chabris, C. F., O’Connor, E., & Breiter, H. C. (2001).

Beautiful Faces Have Variable Reward Value: FMRI and Behavioral Evidence. Neuron,

32(3), 537–551. https://doi.org/10.1016/S0896-6273(01)00491-3

Ariga, A., & Arihara, K. (2017). Visual attention is captured by task-irrelevant faces, but not by

pareidolia faces. 2017 9th International Conference on Knowledge and Smart

Technology (KST), 266–269. https://doi.org/10.1109/KST.2017.7886105

Bakhshi, S., Shamma, D. A., & Gilbert, E. (2014). Faces engage us: Photos with faces attract

more likes and comments on Instagram. Proceedings of the 32nd Annual ACM

Conference on Human Factors in Computing Systems - CHI ’14, 965–974.

https://doi.org/10.1145/2556288.2557403

Black Panther Clipart Black Power—Raised Fist Clip Art—Free Transparent PNG Clipart

Images Download. (n.d.). Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2i8d3N4H7H7G6A0_black-panther-clipart-black-

power-raised-fist-clip-art/
 107
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Bojko, A. (2013). Eye Tracking the User Experience: A Practical Guide to Research. Rosenfeld

Media.

Brown, C. (2017, October 6). Why YouTube Thumbnails Matter. Medium.

https://mag.octoly.com/youtube-thumbnails-matter-563c9af6fa74

Burke, D., Nolan, C., Hayward, W. G., Russell, R., & Sulikowski, D. (2013). Is There an Own-

Race Preference in Attractiveness? Evolutionary Psychology, 11(4),

147470491301100400. https://doi.org/10.1177/147470491301100410

Chudley, J. (2013). A Pocket Guide to Usability of web photos. Five Simple Steps.

fivesimplesteps.com

Clipart—Bible Clipart—Full Size PNG Clipart Images Download. ClipartMax.com. (n.d.).

Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2i8K9G6b1i8G6Z5_clipart-bible-clipart/

Cunningham, C. A., & Egeth, H. E. (2018). The capture of attention by entirely irrelevant

pictures of calorie-dense foods. Psychonomic Bulletin & Review, 25(2), 586–595.

https://doi.org/10.3758/s13423-017-1375-8

Davies, D. (2018, January 7). The 7 Most Popular Search Engines in the World—SEO 101.

Search Engine Journal. https://www.searchenginejournal.com/seo-101/meet-search-

engines/

DeAngelus, M., & Pelz, J. B. (2009). Top-down control of eye movements: Yarbus revisited.

Visual Cognition, 17(6–7), 790–811. https://doi.org/10.1080/13506280902793843

Folk, C. L., Remington, R. W., & Johnston, J. C. (1992). Involuntary covert orienting is

contingent on attentional control settings. Journal of Experimental Psychology. Human

Perception and Performance, 18(4), 1030–1044.

 108
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Giang, V. (n.d.). This Heatmap Proves That Looks Are The Most Important Thing On Your

LinkedIn Profile. Business Insider. Retrieved August 22, 2020, from

https://www.businessinsider.com/should-you-include-a-picture-on-your-linkedin-profile-

2012-5

Grammer, K., & Thornhill, R. (1994). Human (Homo sapiens) facial attractiveness and sexual

selection: The role of symmetry and averageness. Journal of Comparative Psychology

(Washington, D.C.: 1983), 108(3), 233–242. https://doi.org/10.1037/0735-

7036.108.3.233

Henderson, J. M., & Hayes, T. R. (2017). Meaning-based guidance of attention in scenes as

revealed by meaning maps. Nature Human Behaviour, 1(10), 743–747.

https://doi.org/10.1038/s41562-017-0208-0

Hershler, O., & Hochstein, S. (2005). At first sight: A high-level pop out effect for faces. Vision

Research, 45(13), 1707–1724. https://doi.org/10.1016/j.visres.2004.12.021

Heterosis. (2021). In Wikipedia.

https://en.wikipedia.org/w/index.php?title=Heterosis&oldid=1031666455

Homework—Homework Icon Black And White—Full Size PNG Clipart Images Download.

ClipartMax.com. (n.d.). Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2i8K9m2b1H7Z5K9_homework-homework-icon-

black-and-white/

Hosey, C., Vujović, L., St. Thomas, B., Garcia-Gathright, J., & Thom, J. (2019). Just Give Me

What I Want: How People Use and Evaluate Music Search. Proceedings of the 2019 CHI

Conference on Human Factors in Computing Systems - CHI ’19, 1–12.

https://doi.org/10.1145/3290605.3300529
 109
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The Fusiform Face Area: A Module in

Human Extrastriate Cortex Specialized for Face Perception. The Journal of Neuroscience,

17(11), 4302–4311. https://doi.org/10.1523/JNEUROSCI.17-11-04302.1997

Khabab, O. (2020, July 17). Why Video is the King of Content for Marketers—2020 Survey.

https://www.business2community.com/content-marketing/why-video-is-the-king-of-

content-for-marketers-2020-survey-02327948

KLI. (2016, May 11). Eye Tracking: Why Are We Trained to Recognize Other Human Faces?

Key Lime Interactive Blog. https://info.keylimeinteractive.com/eye-tracking-shows-the-

power-of-the-face

Krug, S. (2010). Rocket surgery made easy: The do-it-yourself guide to finding and fixing

usability problems. Berkeley, Calif. : New Riders, c2010.

https://catalog.lib.unc.edu/catalog/UNCb6403880

Krug, S. (2014). Don’t make me think, revisited: A common sense approach to Web usability

(Third edition.). [Berkeley, Calif.] : New Riders, [2014].

https://catalog.lib.unc.edu/catalog/UNCb7904778

Laeng, B., Mathisen, R., & Johnsen, J.-A. (2007). Why do blue-eyed men prefer women with the

same eye color? Behavioral Ecology and Sociology, 61, 371–384.

https://doi.org/10.1007/s00265-006-0266-1

Lewandowski, D., & Kammerer, Y. (2021). Factors influencing viewing behaviour on search

engine results pages: A review of eye-tracking research. Behaviour & Information

Technology, 40(14), 1485–1515. https://doi.org/10.1080/0144929X.2020.1761450

 110
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Lewis, M. B. (2010). Why are mixed-race people perceived as more attractive? Perception,

39(1), 136–138. https://doi.org/10.1068/p6626

Lewis, M. B. (2011). Who is the fairest of them all? Race, attractiveness and skin color sexual

dimorphism. Personality and Individual Differences, 50(2), 159–162.

https://doi.org/10.1016/j.paid.2010.09.018

Lewis, M. B. (2012). A Facial Attractiveness Account of Gender Asymmetries in Interracial

Marriage. PLOS ONE, 7(2), e31703. https://doi.org/10.1371/journal.pone.0031703

Li, Y., & Belkin, N. J. (2008). A faceted approach to conceptualizing tasks in information

seeking. Information Processing & Management, 44(6), 1822–1837.

https://doi.org/10.1016/j.ipm.2008.07.005

Moran, K., & Goray, C. (2019, November 17). Complex Search-Results Pages Change Search

Behavior: The Pinball Pattern. Nielsen Norman Group.

https://www.nngroup.com/articles/pinball-pattern-search-behavior/

Motioncue. (n.d.). The Video Marketing Landscape in 2020. Retrieved September 18, 2020,

from https://motioncue.com/the-video-marketing-landscape-in-2020/

Need A Roommate—Illustration—Free Transparent PNG Clipart Images Download. (n.d.).

Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2H7N4m2H7G6i8d3_need-a-roommate-

illustration/

Oliveira, F. T. P., Aula, A., & Russell, D. M. (2009). Discriminating the relevance of web search

results with measures of pupil size. Proceedings of the SIGCHI Conference on Human

Factors in Computing Systems, 2209–2212. https://doi.org/10.1145/1518701.1519038

 111
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Patel, N. (2016, August 23). 4 Quick Wins to Increase Your YouTube Engagement.

Https://Neilpatel.Com/. https://neilpatel.com/blog/4-quick-wins-to-increase-your-

youtube-engagement/

Pernice, K. (2017, November 12). F-Shaped Pattern of Reading on the Web: Misunderstood, But

Still Relevant (Even on Mobile). Nielsen Norman Group.

https://www.nngroup.com/articles/f-shaped-pattern-reading-web-content/

Rainwater, B. (2013, July 16). Is YouTube Really the Second Largest Search Engine? Pace.

https://www.paceco.com/insights/social-media/youtube-search-engine-myth/

RANDOM.ORG - True Random Number Service. (n.d.). Retrieved November 27, 2017, from

https://www.random.org/

Research Triangle. (2021). In Wikipedia.

https://en.wikipedia.org/w/index.php?title=Research_Triangle&oldid=1022293892

Retrospective Think-Aloud | Qualitative Research | Tobii Pro. (2019, May 27). [Information].

https://www.tobiipro.com/blog/Retrospective-Think-Aloud/

Rhodes, G., Lee, K., Palermo, R., Weiss, M., Yoshikawa, S., Clissa, P., Williams, T., Peters, M.,

Winkler, C., & Jeffery, L. (2005). Attractiveness of Own-Race, Other-Race, and Mixed-

Race Faces. Perception, 34(3), 319–340. https://doi.org/10.1068/p5191

Rhodes, G., Yoshikawa, S., Clark, A., Lee, K., McKay, R., & Akamatsu, S. (2001).

Attractiveness of Facial Averageness and Symmetry in Non-Western Cultures: In Search

of Biologically Based Standards of Beauty. Perception, 30(5), 611–625.

https://doi.org/10.1068/p3123

Riegelsberger, J., & Sasse, M. A. (2001). Trustbuilders and Trustbusters. In B. Schmid, K.

Stanoevska-Slabeva, & V. Tschammer (Eds.), Towards the E-Society: E-Commerce, E-

 112
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Business, and E-Government (pp. 17–30). Springer US. https://doi.org/10.1007/0-306-

47009-8_2

Riegelsberger, J., Sasse, M. A., & McCarthy, J. D. (2003). Shiny happy people building trust?

Photos on e-commerce websites and consumer trust. Proceedings of the SIGCHI

Conference on Human Factors in Computing Systems, 121–128.

https://doi.org/10.1145/642611.642634

Royse, M. (2020, May 12). The rise of pinball searches and what it means for SEO. Knowledge

Enthusiast. https://knowledgeenthusiast.com/2020/05/12/the-rise-of-pinball-searches-

and-what-it-means-for-seo/

Ruel, L. (2007, August 1). OJR’s “five guide” to do-it-yourself website usability testing. Online

Journalism Review. http://www.ojr.org/ojrs-five-guide-to-do-it-yourself-website-

usability-testing/

Sato, S., & Kawahara, J. I. (2015). Attentional capture by completely task-irrelevant faces.

Psychological Research, 79(4), 523–533. https://doi.org/10.1007/s00426-014-0599-8

Stepanova, E. V., & Strube, M. J. (2018). Attractiveness as a Function of Skin Tone and Facial

Features: Evidence from Categorization Studies. The Journal of General Psychology,

145(1), 1–20. https://doi.org/10.1080/00221309.2017.1394811

Shiffman, D. (2015, June 8). How to Create Video Thumbnails That Boost Views

[Www.socialmediaexaminer.com]. Social Media Examiner | Social Media Marketing.

https://www.socialmediaexaminer.com/create-video-thumbnails-that-boost-views/

Shimojo, S., Simion, C., Shimojo, E., & Scheier, C. (2003). Gaze bias both reflects and

influences preference. Nature Neuroscience, 6(12), 1317–1322.

https://doi.org/10.1038/nn1150
 113
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Soda Bottle Vector—Soda Bottle Icon—Free Transparent PNG Clipart Images Download. (n.d.).

Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2i8d3N4b1H7i8G6_soda-bottle-vector-soda-

bottle-icon/

Spangler, T. (2019, May 3). YouTube Now Has 2 Billion Monthly Users, Who Watch 250

Million Hours on TV Screens Daily. Variety.

https://variety.com/2019/digital/news/youtube-2-billion-users-tv-screen-watch-time-

hours-1203204267/

Steinbrück, U., Schaumburg, H., Duda, S., & Krüger, T. (2002). A picture says more than a

thousand words: Photographs as trust builders in e-commerce websites. CHI ’02

Extended Abstracts on Human Factors in Computing Systems, 748–749.

https://doi.org/10.1145/506443.506578

Stress Comments—Stress Symbol—Full Size PNG Clipart Images Download. ClipartMax.com.

(n.d.). Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2i8Z5N4m2i8K9m2_stress-comments-stress-

symbol/

Sugar Clipart Png—Bowl Of Sugar Clipart—Full Size PNG Clipart Images Download.

ClipartMax.com. (n.d.). Retrieved September 28, 2020, from

https://www.clipartmax.com/middle/m2H7K9H7G6K9H7A0_sugar-clipart-png-bowl-of-

sugar-clipart/

Sui, J., & Liu, C. H. (2009). Can beauty be ignored? Effects of facial attractiveness on covert

attention. Psychonomic Bulletin & Review, 16(2), 276–281.

https://doi.org/10.3758/PBR.16.2.276
 114
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Tatler, B. W. (2009). Current understanding of eye guidance. Visual Cognition, 17(6–7), 777–

789. https://doi.org/10.1080/13506280902869213

Theeuwes, J. (1991). Cross-dimensional perceptual selectivity. Perception & Psychophysics,

50(2), 184–193. https://doi.org/10.3758/BF03212219

Theeuwes, J. (2010). Top–down and bottom–up control of visual selection. Acta Psychologica,

135(2), 77–99. https://doi.org/10.1016/j.actpsy.2010.02.006

Thornhill, R., & Gangestad, S. W. (1993). Human facial beauty: Averageness, symmetry, and

parasite resistance. Human Nature (Hawthorne, N.Y.), 4(3), 237–269.

https://doi.org/10.1007/BF02692201

Tullis, T., Siegel, M., & Sun, E. (2009). Are people drawn to faces on webpages? ResearchGate,

4207–4212. https://doi.org/10.1145/1520340.1520641

Valuch, C., Pflüger, L. S., Wallner, B., Laeng, B., & Ansorge, U. (2015). Using eye tracking to

test for individual differences in attention to attractive faces. Frontiers in Psychology, 6.

https://doi.org/10.3389/fpsyg.2015.00042

Wagner, A. (2017, May 15). Are You Maximizing The Use Of Video In Your Content

Marketing Strategy? Forbes.Com.

https://www.forbes.com/sites/forbesagencycouncil/2017/05/15/are-you-maximizing-the-

use-of-video-in-your-content-marketing-strategy/#31be64883584

Weinschenk, S. (2019, January 17). Grabbing Visual Attention With The Visual Cortex.

Smashing Magazine. https://www.smashingmagazine.com/2019/01/grabbing-visual-

attention-with-the-visual-cortex/

Wilson, T. D. (1999). Models in information behaviour research. Journal of Documentation,

55(3), 249–270. https://doi.org/10.1108/EUM0000000007145

 115
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Wolf, A. (n.d.-a). BLACK COMEDY & ACTIVISM | Extra Credit with Dr. Mark Anthony

Neal. Retrieved September 28, 2020, from https://www.youtube.com/watch?v=8-_9-

MHAyCQ&list=PLYZhPUaJyXIbRBZNhBDnB2TC2shDEo5cA&index=4

Wolf, A. (n.d.-b). HOW MUCH HOMEWORK IS TOO MUCH? | Extra Credit with Harris

Cooper. Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=WX2vMCV4cKg&list=PLYZhPUaJyXIbRBZNhBD

nB2TC2shDEo5cA&index=2

Wolf, A. (n.d.-c). IS STRESS ALWAYS BAD? | Extra Credit with Dr. Bridgette Martin Hard.

Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=ucTiRfz6nGw&list=PLYZhPUaJyXIbRBZNhBDnB

2TC2shDEo5cA&index=7

Wolf, A. (n.d.-d). POP CULTURE & THE BIBLE | Extra Credit with Dr. Brent Strawn.

Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=6UyQhmka8BY&list=PLYZhPUaJyXIbRBZNhBDn

B2TC2shDEo5cA&index=3

Wolf, A. (n.d.-e). PUBLIC POLICY & HIGHER EDUCATION | Extra Credit with Deondra

Rose. Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=_IfTCcvgx9M&list=PLYZhPUaJyXIbRBZNhBDnB

2TC2shDEo5cA&index=1

Wolf, A. (n.d.-f). RANDOMLY ASSIGNED ROOMMATES: WHAT ARE THE EFFECTS? |

Extra Credit with Dr. Sarah Gaither. Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=tNHaut29v5U&list=PLYZhPUaJyXIbRBZNhBDnB

2TC2shDEo5cA&index=5
 116
MADE YOU LOOK: DO VIDEO THUMBNAILS WITH PORTRAITS ATTRACT AND
HOLD USERS’ ATTENTION?

Wolf, A. (n.d.-g). WHAT’S THE STORY BEHIND SODA TAXES? | Extra Credit with Kelly

D. Brownell. Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=jR6dS97TrRM&list=PLYZhPUaJyXIbRBZNhBDn

B2TC2shDEo5cA&index=6

Wolf, A. (n.d.-h). WHY DO I CRAVE SUGAR? | Extra Credit with Dr. Diego V. Bohórquez.

Retrieved September 28, 2020, from

https://www.youtube.com/watch?v=6GLA4r7yvPE&list=PLYZhPUaJyXIbRBZNhBDn

B2TC2shDEo5cA&index=8

Yarbus, A. L. (1967a). Eye Movements and Vision. Springer US.

Yarbus, A. L. (1967b). Methods. In A. L. Yarbus (Ed.), Eye Movements and Vision (pp. 5–58).

Springer US. https://doi.org/10.1007/978-1-4899-5379-7_2

zenithmedia.com. (2019, September 16). Online video viewing to reach 100 minutes a day in

2021 |. https://www.zenithmedia.com/online-video-viewing-to-reach-100-minutes-a-day-

in-2021/