ISU580 HW9

1119 – Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., hours spent). I don't see anything on the survey about informed consent? Report does not provide any results (number of subjects, descriptives, correlations). Your 'validity analysis' should have been done by drawing conclusions from a scatterplot showing the degree of association among all data points (not just means). A-

1186 - F

1363 - Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., on hours). I don't see anything about informed consent. Missing composite score (average over all items) and its descriptives. Didn't provide any data or scatterplot to support validity check. A-

1473 – Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., on age and hours). I don't see anything about informed consent. No idea what format your output is in (spv-zip containing a bunch of XML files??). Did not include or summarize results in the narrative report. B

1985 - Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., on age and hours). I don't see anything about informed consent. No narrative report of results. B-

2010 – Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., on age). I don't see anything about informed consent. You were right to exclude demographics from the reliability calc (alpha) - you should only include those items of a composite that will be aggregated together. Did not compute composite score (rather than look at individual items). We hadn't talked about Pearson yet - a scatterplot would have worked well. A-

4382 – Nice survey. Could have gained more information by asking open-ended questions, then quantizing later if needed. I don't see anything about informed consent, demographic questions or a validation measure. Pretty small sample. Did not document any of your results in the report. B

6611 – Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., hours spent). I don't see anything on the survey about informed consent? Report does not provide any results (number of subjects, descriptives, correlations). Your 'validity analysis' could have been done by drawing conclusions from a scatterplot showing the degree of association among all data points (not just means). A-

7975 - Nice survey, but could have gained more information by asking open-ended questions, then quantizing later if needed (e.g., on age and hours). I don't see anything about informed consent. Missing narrative report of results? B-

8017 – Nice survey, but I don't see anything about informed consent (although good that you at least indicated results are anonymous). You might have asked for age in an open-ended way (to use as a demographic) then screen as needed from that. Did not compute a composite value (average) over all items for each subject? The scatterplot should be between this composite measure and some behavioral or criterion measure of addiction (e.g., hours per week playing). B+

8424 – Great that you tried to do test-retest, but should be done with the *same* subjects, and since you should be comparing your composite measure (interval measure) the correct statistic would be Pearson (or actually Intraclass correlation) but since we hadn't covered that yet, a tabular comparison or scatterplot would have worked. Missing all descriptives on your composite (and other) measures. No demographics. No data to back up your stats. B