L6: WILCOXAN-SIGNED RANKS Flashcards
(16 cards)
What is the Wilcoxon Signed-Ranks Test?
A non-parametric alternative to the repeated measures t-test.
A non-parametric alternative to the repeated measures t-test. Doesn’t require what?
It does not require normal distribution of the data.
What kind of data can the Wilcoxon Signed-Ranks Test handle?
Dependent variable can be ordinal or interval/ratio (treated as ordinal).
What is the independent variable like in the Wilcoxon Signed-Ranks Test?
Nominal with one factor and two related levels (e.g., before vs after).
What does the Wilcoxon Signed-Ranks Test (WSRT) compare?
The difference between two treatments on the same group of participants.
How does the WSRT analyze the data?
By ranking the differences between paired scores, ignoring the sign.
What are the assumptions of the WSRT?
Data must be paired and at least ordinal
What happens if there is no difference between treatments?
A: Differences will be small and balanced between positive and negative values.
What indicates a significant difference in the WSRT?
Differences will be larger and mostly positive or mostly negative
What is the dependent variable (DV) in the reaction time study?
Reaction times (means over multiple trials)
What is the independent variable (IV) in this study?
Blood alcohol level (Lower vs Higher)
What is the alternative hypothesis (H1)?
Reaction times are faster in the lower blood alcohol condition than in the higher condition
What is the null hypothesis (H0)?
Reaction times will not differ between low and high blood alcohol levels (alcohol does not affect reaction time).
What is the procedure for calculating the Wilcoxon Signed-Ranks Test (WSRT)?
Subtract scores of one treatment from the other.
Rank all differences by size, ignoring signs.
Sum ranks of positive and negative differences separately.
The smaller sum is the test statistic
Check T against critical values to determine significance
What does a large T value in the Wilcoxon Signed-Ranks Test mean?
It means the positive and negative difference scores are evenly split, so no difference indicating no significant difference between treatments.
What is it when the t value is seen as small indicating a significant result?
Large difference between eh two treatments then most of the difference scores will be associated with one sign
