Data management & analysis Module 9

During Data Collection Use your study cell matrix to keep track of how many subjects you have in each of your 4 cells. Strive to have equal numbers of 10 in each cell. If you used random assignment and one (or more) conditions are closed, take those numbers out of your randomizing procedure (take out the number 3s from your hat). If you used four different links to your survey to achieve random assignment and one of the conditions is filled, remove that link.

In person data collection and data entry In person data Label each questionnaire with Ss number, condition, and date of collection Label all of your variables in an SPSS data file under variable view tab Type in data under data view tab

Online data collection and data entry Online survey administration Download survey responses to Excel and then to SPSS or enter data from survey site by hand to data file Label all of your variables in an SPSS data file under variable view tab Type in data under data view tab

What your data file should like like – it depends on your IVs If you have 2 between subjects IVS, your data file should look like the data set-up from Gaultney’s lab 11 for between-subjects factorial designs on page 89. If you have 1 between subject IV and one within subject IV your data file should look like Gaultney’s lab 11for mixed-factorial designs on page 93. Before you analyze your data, be sure your data file set up resembles Gaultney’s!!

Check your entered data Don’t forget to: Sum responses on items if you are using a total score (by hand or with SPSS) Check for missing data What to do with missing data Reverse code items that need to be reverse coded (by hand or with SPSS) Exclude distractor items that are not being measured

Analyze Your Data Follow Gaultney’s Lab 11 instructions for data analysis exactly for your data. You will then have output of descriptive statistics that show your means for each cell for the interaction, and also the means for each of the IV levels for the main effects (these are labeled “total” you might have to ponder these for a minute to figure out which total belongs to which of the IV levels. If this proves too annoying, you can click analyze-compare means-means-and put your 2 Ivs and 1 DV in the table and analyze – and this output will be easier to interpret! You will have output of your inferential stats - the 3 Fs, one for each IV and one for the interaction.

Gaultney’s Means for Lab 11

Sample write up for Gaultney’s ANOVA In order to test the effects of expecting to drink alcohol and actually drinking alcohol on aggression a 2 (expectation: alcohol vs. water) X 2 (actually drank: alcohol vs. water) between-subjects factorial ANOVA was conducted. This produced a significant main effect of expectation on aggression. That is, participants who expected alcohol reported significantly more aggression (M = 192.4, SD = 31.68) than participants who expected to drink water (M = 164.4, SD = 27.00), F (1, 36) = 8.95, p = .05.

Sample write up continued There was no main effect of drinking on aggression. That is, there were no significant differences in aggression scores between participants who drank alcohol (M = 183.45, SD = 31.68) and those participants who drank water (M = 173.35, SD = 32.98), F (1, 36) = 1.16, p = .22. The results of the interaction analysis showed no significant interaction between participants who expected to drink alcohol and actually drank alcohol (M = 200.4, SD = 30.47), expected to drink alcohol and actually drank water (M = 184.4, SD = 32.37), expected to drink water and drank alcohol (M = 166.5, SD = 23.50), and expected to drink water and drank water (M = 162.3, SD = 31.24), F (1, 36) = .397, p = .53.

Statistics to include in your paper Descriptive Statistics Means and standard deviations for each IV level on the DV separately Means and standard deviations for each of the 4 conditions 2 Graphs for each of your IVs on the DVs (you know how to do this), these will appear in the body of the paper in the results section

Statistics to include in your paper Inferential Statistics Report 3 F statements Post hoc analyses (t-tests if the interaction F was significant) DO NOT put SPSS output in the body of the paper – include it as an appendix