KOFORIDUA TECHNICAL UNIVERSITY

FACULTY OF APPLIED SCIENCE AND TECHNOLOGY

DEPARTMENT OF COMPUTER SCIENCE

BSc. BUSINESS INFORMATION SYSTEMS

MIGBODZI JOSEPH

INDEX: B207230014

RESEARCH METHODOLOGY
Assignment 1

Exercise 1

a. Systematic Bias

i. Explanation of Systematic Bias:

Systematic bias occurs when there are consistent, non-random errors in data

collection or analysis that lead to incorrect conclusions. Unlike random errors, which

can cancel out over a large sample, systematic biases consistently skew the results in a

specific direction, potentially compromising the validity of a study. This bias can

affect any stage of the research process, including sampling, data collection,

measurement, and analysis, leading to results that are not representative of the actual

situation or population.

ii. Causes of Systematic Bias:

Sampling Method Errors: Systematic bias can occur if the sampling method does not

accurately represent the population. For example, if a survey is conducted only among

a specific group (e.g., urban residents) but is supposed to represent a larger population

(e.g., the entire country), the results will be skewed. Non-random sampling

techniques, such as convenience sampling, can lead to systematic bias because they

do not give every individual in the population an equal chance of being selected.

2. Measurement Errors:

Measurement errors arise when the instruments or procedures used to measure

variables consistently produce inaccurate results. For instance, if a scale used to weigh

1
participants is consistently off by a certain amount, all measurements will be

systematically biased. This type of error can be due to faulty equipment, improper

calibration, or incorrect measurement procedures.

3. Data Collection Errors:

Data collection errors occur when the methods used to gather data introduce bias. This

can happen through leading questions in surveys, where the wording of the question

influences respondents' answers. For example, asking "Don't you think that X is the

best option?" can lead to a bias towards agreement. Additionally, interviewer bias,

where the interviewer's tone or behavior influences responses, can also lead to

systematic bias.

b. Research and Sample Design

i. Research:

Research is a structured and systematic process of investigating and studying

materials, sources, or situations to establish facts and reach new conclusions. It

involves identifying a problem or question, formulating a hypothesis, collecting and

analyzing data, and then interpreting the results to contribute to knowledge in a

particular field. Research can be qualitative, focusing on understanding phenomena,

or quantitative, focusing on measuring and analyzing variables.

ii. Sample Design:

Sample design refers to the strategy or plan used to select a subset (sample) of

individuals or items from a larger population for study. The design of the sample has a

significant impact on the reliability and generalizability of the study's findings. A

2
well-designed sample should be representative of the population, ensuring that the

results can be generalized. There are several sampling methods, including random

sampling, stratified sampling, and cluster sampling, each with its advantages and

limitations depending on the research objectives.

Exercise 2

a. Three Formats of Data in Analytics Research:

1. Structured Data:

This is highly organized and formatted in a way that is easily searchable in

databases, such as rows and columns in spreadsheets or relational databases.

Examples include numerical data, dates, and categorical data.

2. Unstructured Data:

This type of data does not have a pre-defined structure and is often more complex to

process. Examples include text from social media posts, videos, images, and audio

recordings. Analyzing unstructured data often requires advanced techniques like

natural language processing or machine learning.

3. Semi-Structured Data:

Semi-structured data has elements of both structured and unstructured data. It does

not conform to a strict structure but contains tags or markers that separate elements,

making it easier to analyze than purely unstructured data. Examples include JSON or

XML files, which have a flexible structure but still contain identifiable markers.

3
b. General Analytics Techniques (GAT) Karen Can Use:

To analyze sensor data and clickstream data, Karen can employ the following four

General Analytics Techniques:

1. Descriptive Analytics: Summarizes and interprets historical data to identify

trends and patterns. This can help Karen understand the overall behavior reflected in

the data.

2. Diagnostic Analytics: Looks into past data to determine why something

happened. Karen can use this to identify the causes behind observed behaviors in the

data.

3. Predictive Analytics: Uses statistical models and machine learning algorithms to

predict future outcomes based on current and historical data. This can help forecast

future user behaviors.

4. Prescriptive Analytics: Recommends actions based on the analysis of data.

Karen can use this to suggest changes or improvements to influence future behaviors.

c. Six Critical Criteria for Mixed-Methods Research Data Analysis:

1. Integration of Data: Ensuring that qualitative and quantitative data are effectively

combined to provide comprehensive insights.

2. Rigor:Maintaining strict standards in data collection, analysis, and interpretation

to ensure validity and reliability.

3. Consistency: Ensuring that the methods and analysis are consistent across the

different types of data.

4. Ethical Considerations: Addressing ethical issues related to data collection and

analysis, such as confidentiality and informed consent.

4
 5. Validity: Ensuring that the analysis accurately reflects the data and research

objectives.

6. Interpretation of Results: Drawing meaningful and actionable conclusions that are

consistent with the mixed-methods approach.

Exercise 3

a. Hypothesis Testing Procedure Outline:

1. Formulate Hypotheses: Start with a null hypothesis (H0) and an alternative

hypothesis (H1).

2.Choose the Significance Level:Determine the alpha level (commonly 0.05).

3. Select the Test Statistic: Depending on the data type, choose an appropriate

statistical test (e.g., t-test, ANOVA).

4. Compute the Test Statistic:Use the data to calculate the value of the test statistic.

5. Make a Decision: Compare the calculated statistic with the critical value to decide

whether to reject or fail to reject the null hypothesis.

6. Interpret the Results: Draw conclusions based on the decision.

b. Differences Between Qualitative and Quantitative Data Analysis:**

Qualitative Data Analysis: Focuses on interpreting non-numerical data to understand

concepts, opinions, or experiences. It involves identifying patterns, themes, and

meanings within the data.

Quantitative Data Analysis: Involves statistical analysis of numerical data to quantify

relationships, test hypotheses, and make predictions. It is more structured and

typically uses software for data analysis.

5
c. Strengths and Weaknesses of Quantitative Research:

Strengths:

1. Objectivity: Results are less prone to personal biases.

2.Reproducibility:Methods and results can be replicated in other studies.

3. Generalizability: Large sample sizes allow findings to be generalized to the broader

population.

Weaknesses:

1.Limited Depth: It may not capture the nuances of complex phenomena.

2. Context Ignorance: Quantitative research often ignores the context or environment

in which data is collected.

d. Threats in Mixed-Method Research for Kofi:

i. Research Design:

Potential threats include poorly defined research questions that do not align well with

the mixed-methods approach, and lack of integration between qualitative and

quantitative components.

i. Data Collection:

Threats could include bias in qualitative data collection, such as interviewer bias, and

sampling errors in quantitative data collection.

iii. Data Analysis and Interpretation of Results:

Challenges include integrating the two types of data in a meaningful way and

avoiding the overemphasis of one type of data over the other.