Exercise: A series of items is presented that may represent a series subscales, or maybe 1 big scale.
Exercise:
A series of items is presented that may represent a series subscales, or maybe 1 big scale. Your broad task is to determine how many factors these items represent. The items are attitudes and opinions regarding condom use. College students formed the base sample.
Your tasks:
- Using an 'eyeball' approach, try and 'group' the items the best you can - how many factors can you form that 'make sense' based on simply looking at the variable descriptions? Name the factors.
- In SPSS run a PFA (not the default PCA) to determine the number of factors. Evaluate the scree plot to determine the number of plausible factors. How many factors should be derived based on evaluating the scree plot? Do NOT name the factors here or rotate.
- In SPSS run a PFA and force a 2-factor solution. Rotate your findings using a Varimax rotation. Name the factors.
Product: Use SPSS to test univariate and multivariate assumptions of factor analysis - try and cover most of these and do a quick bullet-point assessment. It does NOT need to be as detailed as our other exercises.
Do a typical writeup first outlining the goal of the exercise, then covering assumptions, then address findings. From your eyeball approach (#1), how should the items be placed together? How many factors? And from #1, name these factors. From #2, how many factors? From #3, you force a 2 -factor solution and rotated the findings - how do these findings compare to #1? How would you name these two factors? Create a table of your findings from #3 (the rotated solution) per T&F examples, and blank out factor loadings under .30. Summarize your overall findings from #3 to end the writeup.
Deliverable: Word Document
