L19: CHOOSING THE CORRECT STASTICAL TEST

Question 1

What test for one group against a value?

Answer 1

One sample t-test

Question 2

What test for comparing two conditions?

Answer 2

Paired t-test
Independent samples t-test
Mann-Whitney U
Wilcoxon signed ranks

Question 3

Whats the test comparing more than two conditions?

Answer 3

One way ANOVA
Repeated measures ANOVA
Kruskall-Wallis
Friedman’s test

Question 4

Describe one sample t-test?

Answer 4

Used when you have one group of partipants and need to compare their mean against a specific value

Question 5

Whats the hypothesis for one-sample t-test?

Answer 5

• Null hypothesis: Average score will not differ from four
• Alternative hypothesis: Average score will be higher than four
• Dependent variable: Enjoyment rating (interval)

Question 6

Describe the hypothesis for t-tests that compare the means of two groups/conditions?

Answer 6

Hypothesis H1– there are differences in
the DV as a function of the IV (top figure)
• Null H0 – no difference – IV does not
significantly influence DV (bottom figure

Question 7

What t-test for between subject design?

Answer 7

Independent t-test

Question 8

Whats the t-test for within subject designs?

Answer 8

Paired t-test

Question 9

Whats the assumptions for t-tests?

Answer 9

Normality
Equality of variance
No outliers
Interval or ratio data

Question 10

Explain independent t-test?

Answer 10

Comparing means of 2 separate groups to find if they’re significantly different
Between subject design

Question 11

Whats the variables for an independent t-test?

Answer 11

Independent variable - group 2 levels (autistic vs neurotypical)
Dependent variable - reaction time (ratio)

Question 12

When should you not use an independent t-test?

Answer 12

If the Shapiro-Wilk test is significant (p < .05) then your data is not
normally distributed and transformation of the data or using non-
parametric tests (Mann Whitney U test) should be considered instead
of the independent T test

Question 13

Describe paired t-test?

Answer 13

Within-subject designs where same participants are tested under two conditions

Question 14

In the example study on self-affirmation and spelling, what is the independent variable?

Answer 14

Affirmation with 2 levels: low (before affirmation) and high (after affirmation).

Question 15

What is the dependent variable in the self-affirmation spelling test study?

Answer 15

Spelling test scores (ratio level).

Question 16

What are the null and alternative hypotheses in this t-test example?

Answer 16

Null hypothesis: No difference in spelling scores between low and high affirmation conditions.
Alternative hypothesis: Spelling scores are higher in the high affirmation condition.

Question 17

How to interpret the effect size of cohens d

Answer 17

0.2 - small
0.5 - medium
0.8 - large

Question 18

Which ones are the non-parametric alternatives to t-test?

Answer 18

• Mann-Whitney - between subject design
• wilcoxon - within subjects design

Question 19

What type of design is the Mann-Whitney U Test used for?

Answer 19

Independent samples design, comparing two groups on one measure

Question 20

Does the Mann-Whitney U Test use means to compare groups?

Answer 20

No, it uses ranks (order of scores), not means.

Question 21

What indicates a significant difference in the Mann-Whitney U Test?

Answer 21

The ranks in one condition are systematically higher or lower than in the other.

Question 22

What pattern of scores indicates no difference between groups?

Answer 22

Large and small scores are mixed evenly across both samples

Question 23

What do rank patterns in the Mann-Whitney U test indicate about differences between two samples?

Answer 23

(a) If scores from the two samples cluster at opposite ends of the rank order, this suggests a systematic difference between treatments.
(b) If scores from the two samples are intermixed evenly along the scale, this indicates no consistent difference between treatments

Question 24

In the wire loop game study using a Mann-Whitney U test, what are the key elements?

Answer 24

Independent variable: Type of drink (alcoholic vs. non-alcoholic)
Dependent variable: Position on the wire loop game (ordinal ranking)
Null hypothesis: Type of drink does not affect game performance.
Alternative hypothesis: Non-alcoholic drink consumers will perform better (get further) than alcoholic drink consumers

Question 25

What is the Wilcoxon Signed-Ranks Test used for and how does it work?

Answer 25

Used in within-subjects designs to compare two treatments on the same participants. It ranks the differences between paired scores (ignoring sign). If no difference, differences are small and mixed positive/negative. If a real difference exists, differences are larger and mostly positive or mostly negative.

Question 26

What are ANOVAs used for, and what are the key types?

Answer 26

Compare means between three or more groups. One-way ANOVA: Between-subjects design (different participants in each group). Repeated measures ANOVA: Within-subjects design (same participants across conditions

Question 27

What are the main assumptions of ANOVA?

Answer 27

Dependent variable is interval or ratio scale. Homogeneity of variance (equal variances across groups). Dependent variable is normally distributed within groups.

Question 28

In the One-Way ANOVA study on birth season and phonics scores, what are the key elements?

Answer 28

Design: Between-subjects (different children in each birth season group) Independent variable: Birth season with 4 levels (Autumn, Winter, Spring, Summer) Dependent variable: Phonics screening score (ratio scale) Null hypothesis: No effect of birth season on phonics scores Alternative hypothesis: Birth season affects phonics scores

Question 29

What does Levene’s test indicate in One-Way ANOVA, and what should its result be?

Answer 29

Levene’s test checks homogeneity of variance (equal variances across groups). It should be non-significant to meet the assumption, meaning variances are equal.

Question 30

What is the assumption about data distribution in One-Way ANOVA?

Answer 30

The dependent variable should be normally distributed within each group.

Question 31

: In the repeated measures ANOVA study on information retention, what are the key components?

Answer 31

Design: Within-subjects (same participants experience all three conditions) Independent variable: Task type with 3 levels: Encoding only (repeated reading) Encoding + retrieving (reading + recall practice) Control (separate task) Dependent variable: Number of words recalled (ratio scale) Null hypothesis: No difference in recall between the three task types Alternative hypothesis: At least one task type leads to different recall performance

Question 32

What does a non-significant test of sphericity indicate in repeated measures ANOVA?

Answer 32

It means the assumption of sphericity is met, so no correction to degrees of freedom is needed.

Question 33

How do you interpret the following ANOVA result: F(2, 198) = 52.36, p < .001, η² = .346?

Answer 33

There is a significant effect of task type on recall. The effect size (η² = .346) shows a large effect.

Question 34

What does post-hoc testing show in the example?

Answer 34

More words recalled in encoding + retrieving than in both encoding only and control (p < .001). More words recalled in encoding only than in control (p < .001).

Question 35

What is the Kruskal-Wallis test used for?

Answer 35

It compares 3 or more independent groups on an ordinal or non-normally distributed variable. It is a non-parametric alternative to One-Way ANOVA. Similar to the Mann-Whitney U test, which compares only 2 groups

Question 36

What do rank patterns in the Kruskal-Wallis test indicate?

Answer 36

(a) If scores from the three groups cluster in specific rank areas, this suggests a systematic difference between treatments. (b) If scores from the three groups are evenly intermixed along the ranks, this indicates no consistent difference between treatments.

Question 37

In the Kruskal-Wallis study on music tempo and duration estimation, what are the key elements?

Answer 37

Design: Independent groups with 3 levels of tempo (fast, medium, slow) Independent variable: Tempo (3 levels) Dependent variable: Duration estimation Hypotheses: Null: Tempo has no effect on duration estimation Alternative: Tempo affects duration estimation Reason for using Kruskal-Wallis: Data violated assumptions for One-Way ANOVA, so a non-parametric test was used.

Question 38

What is Friedman’s test and when is it used?

Answer 38

It’s a within-subjects non-parametric test, used as an alternative to repeated measures ANOVA. It ranks the conditions within each participant. Used when comparing three or more related conditions with data that don’t meet parametric assumptions.

Question 39

In the Friedman’s test study on music type and maths performance, what are the key details?

Answer 39

Design: Within-subjects (same participants tested under all music conditions) Independent variable: Music type with 3 levels (no music, dance music, classical music) Dependent variable: Maths test score Null hypothesis: Music type has no effect on maths performance Alternative hypothesis: Music type affects maths performance Reason for using Friedman’s test: Data violated assumptions of repeated measures ANOVA, so a non-parametric test was needed.

Question 40

Practice MCQ • A study is comparing the drawing skills of left-handed and right- handed children. First the researchers determined whether each child was left-handed or right-handed. They then measured each child's accuracy in copying a picture of a castle. The data was found to NOT meet parametric assumptions. What would be an appropriate statistical test to use to analyse the data? a) Paired sample t-test b) Independent t-test c) Mann-Whitney U d) Wilcoxon

Answer 40

Mann-Whitney U