Correlation Review 1 Write in APA style the

Correlation Review 1. Write in APA style the results of each of the 2 major correlations on the handout provided: – Policy-holder age and number of claims – Policy-holder age and cost of claims

Possible Write-Up n There was a significant correlation between age and average cost of claim, such that younger drivers had more expensive claims, r (123) = -. 235, p <. 01. n There was a significant correlation between age and number of claims, such that older drivers had more claims, r (128) =. 358, p <. 001.

Experimental Designs

Research Design n The way that you select and assign participants to groups and control for extraneous variables.

Two Group Design n Simplest Experimental Design One IV (two levels) Assignment to groups is either: – Between – Within

Assignment to Groups Random vs. Correlated

1. Random Assignment n Equal chance of being assigned to one of all possible groups. i. e. , Independent Groups i. e. , Between-Subjects Comparison

2. Correlated Assignment Participants are related in some way; a small number of paired participants are randomly assigned to the treatment conditions. n i. e. , Dependent Groups or Within. Subjects Comparison

Types of Correlated Assignment

1. Matched Pairs Type of correlated assignment in which subjects are first measured on some variable and then are paired with someone who scored the same or similarly on that variable. Then the members of the pair are randomly assigned to one of the two levels.

2. Natural Pairs Type of correlated assignment in which naturally occurring pairs are used. e. g. , Twins, married couples, roommates, etc.

3. Repeated Measures Type of correlated assignment in which the same participants are used in all conditions.

Why Use Correlated Assignment? n Small number of participants – May yield unequal groups on important variables. n n Allows us to control for major extraneous variables. In the case of repeated measures we control for all individual differences. Statistical Advantage

Why Use Random Assignment? n n Controls for lots of individual and external EV if you have lots of subjects. Controls for unknown EVs. Easier to do. May be no choice due to the nature of the study or participants.

Comparing Groups Understanding and Using the t-test

t-test Evaluates the difference between two means. Null hypothesis: The groups means are equal (any difference is due to chance variation only) t = Difference Between Groups (i. e. , IV) Differences Within Groups (i. e. , Error)

Independent Samples t-test n Use when two randomly assigned group means are being compared. n Assumptions about your data that must be met: – Independent samples – Normally distributed – The samples have the same variance (or close to the same amount of variability)

A Demonstration n DV: Scores on the Survey of Recent Life Experiences (measure of hassles) – “Dislike daily activities” – “Conflicts at work” – “Having your trust betrayed” n Scores can range from 41 -164

Independent Groups n n Capital Citizens District 12 Citizens

Enter data into Graphpad. com District 12 12 150 72 142 81 160 69 149 44 130 61 146 55 152 43 148 60 121 52 136

Components n n n n N Means and Standard deviations Standard Error of the difference 95% Confidence interval DF t-value p-value

Writing up the independent samples t-test in APA style. There was a significant difference between citizens from the Capital and citizens from D 12 in the amount of hassles experienced on a regular basis, t(18) = 12. 47, p <. 001. Those from D 12( m = 143. 4, sd = 11. 5) had significantly more hassles than those from the Capital (m = 54. 9 , sd = 19. 2). Enter the df here. Enter the t-value here.

Writing Non-Significant results While Group A scored higher (m = 4, sd = 2) than Group B (m = 3. 5, sd = 1. 4) on the memory test, the difference was not statistically significant, t(14) =. 96, p =. 35).

In-Class Work n Practice conducting and interpreting a ttest.