PSYC*1010 Chapter 11: t Tests for Two Related Samples Flashcards
When is a repeated-measures t-test conducted?
When two sets of data from the same group are collected and compared
In a repeated-measures test, which variable is measured two or more times for each individual in a single sample?
The dependent variable
What is the main advantage to using a repeated-measures method?
The is no risk that participants in one treatment are substantially different from the participants in another (because the same individuals are used in all treatment conditions)
What is the main disadvantage to using a repeated-measures method?
Design allows for factors other than the treatment effect to change a participant’s score (ex. time related factors)
What is a matched-subjects design?
Involves two separate samples, but each individual in one sample is matched one-to-one with an individual in the other sample
How are participants matched in a matched-subjects design?
Based on one or more variable considered to be especially important for the study
The closest to perfect a match can be in a matched-subjects design is limited by what?
The closest to perfect a match can be is limited to the variables used for matching
What is the common name that refers to both repeated-measures and matched-subjects designs?
Related-samples design
How are difference scores typically obtained?
By subtracting the first score (before treatment) from the second score (after treatment)
What does the sign of each difference score indicate?
The direction of change
For a repeated-measures design, what serves as the sample data for the hypothesis test?
The sample of difference scores
T or F: A selected sample from a population likely have error, so even if μD=0, MD is not expected to be exactly zero.
True
In a repeated-measures t-test, what is population mean always set to?
Zero
What must be calculated before computing estimated standard error in a repeated-measures test?
The variance/standard deviation using the sum of squares from the sample of difference scores (calculated using a table)
What columns are included in the table used to calculate sum of squares from the sample of difference scores?
X1, X2, D, D^2
What are the two basic assumptions of a repeated-measures t-test?
- Observations within each treatment condition must be independent
- The population distribution of difference scores must be normal (or n>30)
What are two factors that affect confidence intervals?
- Percentage of confidence used
- Sample size
How does percentage of confidence affect the size of a confidence interval?
Larger percentage produces a wider interval
Why are confidence intervals not a pure, unbiased measure of effect size?
Because they’re influenced by sample size
Is standard error directly or inversely related to sample size?
Inversely
Is standard error directly or inversely related to sample variance?
Directly
T or F: Sample size influences measures of effect size.
False
T or F: Variance influences measures of effect size.
True
Does a large variance produce a small or large value for the t statistic?
Larger variance produces a smaller value for the t statistic
Does an independent-measures or repeated-measures design typically require fewer subjects?
Repeated-measures
Does an independent-measures or repeated-measures design use participants more efficiently?
Repeated-measures
Are independent-measures or repeated-measures designs more suited for studying changes that take place over time?
Repeated-measures
What are individual differences?
Characteristics that vary from one individual to another
What are time-related factors?
Factors other than the treatment effect that cause a participant’s score to change (ex. participant’s mood, weather, etc…)
What are order effects?
Changes in scores that are caused by participation in an earlier treatment (ex. participant gaining experience, and thus performing better)
What is one strategy for dealing with time-related and order effects?
Counterbalancing
What does counterbalancing involve?
One group receives treatment 1 followed by treatment 2, and the other group receives treatment 2 followed by treatment 1
How are degrees of freedom calculated to find the critical t value for a related-measures t-test?
df (sample 1) + df (sample 2)
