Ch. 10 Exam (Exam 4)

What are the assumptions of a dependent-group samples t-test?

1. IV (or predictor) is dichotomous (nominal scale w/ 2 groups) 2. Groups are dependent via matching or repeated measures 3. n of 2 groups is equal 4. DV (or outcome) is interval or ratio 5. DV (or outcome) is normally distributed

Concerns for Repeated measures design (7) (COMP FC)

1. Order effect 2. Carryover 3. Practice 4. Fatigue 5. Maturation/history 6. Takes time and effort 7. Can't change participant in a permanent way

Fatigue effect

A confound that occurs when changes in the DV occur because of participants becoming tired

Carryover effect

A confound that occurs when the effect of one condition of the treatment continues (or carries over) into the next condition.

Order effect

A confound that occurs when the order of each treatment condition cannot be separated from the condition

Dependent-groups design

A design where participants in different conditions are related or are the same person - More sensitive & More powerful than independent-groups design - Decreases random error from individuals -Increases power, which is the ability to detect the difference

Cohen's d

A measurement of effect size; describes the magnitude of the effect of our IV (or predictor) on the DV (or outcome) in standard deviation units.

Counterbalancing

A procedure to remove order effects in a repeated measures experiment or quasi-experiment Participant's experience the different levels of the IV in different orders randomly assigned to the different orders even numbers of participants assigned to the different orders

Latin square balancing

A type of partial counterbalancing where each condition appears once in each sequence; participants are then randomly assigned to different sequences sudoku

Repeated measures design is also called

Between-subject design

Which of the following is NOT an assumption of the dependent-samples t test? A. Predictor is dichotomous B. The n of the groups is equal C. The groups are independent D. The outcome is interval or ratio

C. The groups are independent

____________ is an appropriate measure of effect size for dependent-samples groups

Cohen's d

Practice effect

Confound that occurs when the participants' score changes due to repeating a task rather than because of the level of the IV

Power of a repeated-measures dependent design comes from

Controlling all potential confounding participant variables

Where does the power of a repeated measures design come from?

Controlling all potential confounding participant variables

Dependent-groups design decreases _______, which is created from __________

Decreases "random error", which is created from participant characteristics This will decrease the chances of confounds due to participant variables and provides assurance of homogeneity of variance This is because the participants show similar characteristics in each group

Repeated measures design (within-subjects design)

Design in which participants experience every condition in a study; in an experiment, they also randomly assigned to the order of conditions.

Matched-pairs design

Design in which participants in each group are matched on a characteristic relevant to the variable that is being measured; in an experimental design, a member of each matched pair is randomly assigned to each IV condition.

Matched groups two-group design and repeated measures two-group designs are analyzes with different types of t tests. True or false?

False

Standard error of the mean difference (SD D)

Standard deviation of the difference for a sampling distribution of mean differences; estimated from the standard deviation of the difference scores in a dependent-samples study

Dependent-samples t-test (or paired-samples t-test/within-subjects t-test)

Statistical test used to analyze results from a dependent two-groups design

Appropriate statistic for a simple experiment

independent samples t test

Appropriate statistic for a multiple-groups design

one-way ANOVA

Appropriate statistic for a matched-pairs design

paired-samples t test

Appropriate statistic for a repeated measured design

paired-samples t test

Appropriate statistic for a repeated measured design dependent multiple-groups design

repeated measures ANOVA

Partial Eta Squared

the effect size for a dependent multiple-group design that removes the variability unique to individual participants (SS s) from the error term

Overall dependent-groups designs _______power of a study, because you ___________ the difference between groups and ________the error of variability in scores

- Increase - Maximize - Decrease

Why is it necessary in a repeated measures design?

. Counterbalancing is necessary in a repeated measures design, because it avoids confounding orders and condition

Concerns for Matched-Pairs Design (4) (CAAM)

1. Can't get perfect match 2. Attrition double (lose one person, lose pair) 3. Additional screening / testing of participant (e.g., participants know purpose study; SENTIZATION, fatigue, demanding more from participants) 4. More time and effort

Assumptions for one-way within-subjects ANOVA

1. DV is interval or ratio 2. DV is normally distributed 3. Sphericity in variances of differences between pairs of groups 4. IV has 3+ levels 5. Groups are dependent

Partial counterbalancing

Randomly assigning participants to different sequences of conditions so that each condition is represented in each order an equal number of times, not all sequences are represented when participants are randomly assigned to different sequences of conditions so that each condition is represented in each order an equal number of times but not all sequences are represented

Overall dependent groups designs

Increase power of study Maximize difference between groups Minimize error variability in scores

What is the major advantage of dependent designs?

Increases power - ability to detect the difference Decreases the random error from individual participants More sensitive

Which of the following is NOT a characteristic of a dependent-groups design when compared to an independent-groups design: A. Assurance of homogeneity of variance B. Decreased chance of confounds C. Greater sensitivity to changes in the measured variable D. Larger sample size

Larger sample size

Advantage of matched pairs design

Minimizes error due to individual differences Matching ensures groups are equivalent Reduce number of participants needed Increases number of observations

Randomized partial counterbalancing

Randomly assigning each participant to one of the possible sequences of conditions w/out concern about order or sequence; used when you have a larger number of sequences than participants

Complete counterbalancing

Randomly assigning participants to all the possible sequences of conditions in an experiment

Advantage of repeated measures design

Reduces error variance Increased power Decrease error variance

Sphericity

The assumption that the variances of the differences between all the combinations of pairs of groups are equal

Mean Difference (M D)

The average difference between the scores of the matched pairs or the scores for the same participants across two conditions

Standard error of the difference between means

The average variability in a sampling distribution of differences between means.

Where does the power of the dependent groups design come from?

The decrease in random error that is created by participant characteristics

Where does the power of the dependent-groups design come from

The decrease in random error that is created by participant characteristics

Within-subjects ANOVA (or repeated measures ANOVA/dependent-groups, one-way ANOVA)

The statistical test used to analyze dependent multiple-groups design

Matched pair design is also called

Within-subject design