ISU580 HW10

1119 – Awesome. A

1186 -

1363 -

1473 – Part 1: setup looks good, but the df for chi-square is 4 - should be 2 for 3 categories - so I don't quite understand how you set up the calculation. Also, the output shows p=.132 which is not significant, contrary to your results. Part 2: Great, but you should have stopped at the correlation - I don't understand the t-test. A-

1985 -

2010 – Awesome. A

4382 – Part 1: your reading of the SPSS results is correct, but I don't think I agree with the setup for the analysis - should just be looking at a single nominal variable, you imply that you are looking at several. Part 2: Nice correlations and scatter plots, but only needed to do one - between your composite measure and a validation measure. A-

6611 –

7975 - Comparison on Chi-square is a population with equal frequency distributions across your 3 categories. Should have done the Chi-square on your aggregate composite measure, rather than individual items. There should only be 3 observed/expected frequency terms (and you could have let SPSS compute it for you). Part II is missing narrative description and conclusions - but again this should have just been a single Pearson calculation in SPSS. B-

8017 – Part 1: your narrative looks good, but the chi-square test itself shows a df of 3, indicating you used 4 categories for testing, yet your narrative indicates you used 3? Part 2: should do a Pearson correlation, not a regression, and 0.7 is not an absolute threshold for significance (depends, in part, on the number of samples). If you had run SPSS correlation it would have given you the signficance level. B

8424 – Instructions were to use 3 categories, this would have given you more of a chance of finding a significant difference. However, with the 5 categories you describe, your df should be 4, yet the SPSS output indicates your df is 8??? Assuming the chi-square calculation were corrct, your conclusion (My sample didn’t fall well into the 3 categories.) is backwards - a non-significant (p>.05) result would indicate that your sample did not differ significantly from an equal distribution across N categories (although here you are talking about 3 categories, as opposed to the 5 or 9 referenced earlier??). Part 2 was to do a Pearson correlation, not a chi-square, and you did not provide enough information about what you were testing. C