Scribd
Upload
204 views19 pages

SPSS Summary: Correlation + Regression + ANOVA + T-Test

This document provides an overview of different statistical analyses that can be conducted using SPSS, including correlation, regression, ANOVA, and t-tests. It explains the assumptions, procedures, and reporting for each type of analysis. Key points covered include the different types of variables, when to use correlation versus regression, and the steps and values to report for regression analysis, two-way ANOVA, and independent t-tests. Examples are provided to demonstrate how to interpret and report results for each statistical technique.

Uploaded by

Carson Yu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
204 views19 pages

SPSS Summary: Correlation + Regression + ANOVA + T-Test

This document provides an overview of different statistical analyses that can be conducted using SPSS, including correlation, regression, ANOVA, and t-tests. It explains the assumptions, procedures, and reporting for each type of analysis. Key points covered include the different types of variables, when to use correlation versus regression, and the steps and values to report for regression analysis, two-way ANOVA, and independent t-tests. Examples are provided to demonstrate how to interpret and report results for each statistical technique.

Uploaded by

Carson Yu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

SPSS

Summary
Correlation + Regression + ANOVA + T-test

戴頭盔: For any details, please refer to Yan’s lecture notes!


Types of data
• Nominal: Categorical nature
• e.g., gender
• Ordinal: The position in a serial order
• e.g., champion and the first runner-up in a race
• Interval: The degree between any equally separated countable values
are identical
• e.g., ʻ1ʼ, ʻ2ʼ and ʻ3ʼ
• Ratio: most of the measurements in science are in ratio scale,
• e.g., Kg, cm, L
Test of Mean Differences
• One sample t-test (a score compared with the standard)
• Paired sample t-test (pre-post comparison)
• Independent sample t-test (comparison of two means)
All t-test involve 1 IV and 1 DV, DV is scale/ ratio; IV is nominal (groups)
Test of Mean Differences
• One Way ANOVA (1 IV with many levels, 1DV)
• Two Way ANOVA (2 IV with many levels, 1DV)
• Within Subject ANOVA (repeated measure ANOVA)
• (Comparison of multiple time points)
Test of relationship
• Correlation (Pearson’r)
• Regression
• Cronbach Alpha (reliability)→Questionnaires
Correlation, regression

Decision
Case No A B C D • No IV/DV → correlation (Pearson’ r)
1 10 12 18 26
2 20 14 20 28
• IVs/ DV → regression (only 1 DV)
3 30 16 22 30
4 40 18 24 32
Correlation
• r(n)= … p<.05, <.01, <.001 , or >.05
• Effect size = r square = % of variance explained.
Example:
r(10) = -.631, p=0.05, yields marginal significance.
r square =0.3981, which explains 39.81% of variance.
Regression : steps
• 1. Read “ANOVA” table (report F(df1, df2), p<.05……etc.)
• 2. If ANOVA table is significant, then read the following index:
• b, beta, t, p
• Report those factors (IV) that are significant
• ALSO, report those factors (IV) that are NOT significant
• Beta is the effect size for individual factors
• OVERALL ADJUSTED R Square (capital letter R) – overall variance explained by
significant IVs.
Example 1:
Df1, df2 = F(4,5) =2.465, p>.05 , the model is not
significant.
Example 2:
F(4,5) = 12.465, p<.05, overall model is significant.
The following IVs can significantly predict DV:
IV1 (B=-0.557, Beta = -0.668, t=-1.890, p<.05),
IV3 (B=0.238, Beta= 0.341, t=1.177, p<.05).
IV1 negatively predicts DV, while IV3 positively predicts DV.
Beta of IV =-.668 indicates that when IV1 increase for 1SD, DV
will decrease for .668 SD.
Last Step:
Adjusted R square =0.394, indicating that
significant IVs can predict 39.4% of variance in DV.
Two-way ANOVA
• Assumption test: homogeneity test (in option) > equal variance
assumption.
• Levene’s Test F(df1,df2) = ……, p>.05
• Main test (option: effect size, power, descriptive)
• Report F (df1, df2) = , p<.05…
• Effect size = partial eta square, (0.01, 0.06, 0.14)
• Post-hoc Tukey test: report significant mean differences
• Mean difference, p-value (significant case only)
• Plot graph (smaller variable : separate lines)
F (1,6) = 24.00, p<.01, partial eta squared =
0.80, which is very high.
From post-hoc Tukey test, Group 1 is significantly
smaller than Group 3 (mean difference =-5,
p<.01). Other group comparisons do not yield
statistical significance.
Independent sample t-test
• Assumption test: Levene’s test (report F, p >.05 (good))
• Main test : t(df) = … p<.05, <.01, <.001
• Which group has larger mean?
From Levene’s Test, F=0.000, p>.05, equal
variance is assumed, t(10)=-5.56, p<.001.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy