Pasw spss ver 18

Data were downloaded from SurveyMonkey into Excel (Microsoft Corp., Redmond, WA, USA) and loaded in PASW SPSS ver. 18.0 (SPSS Inc., Chicago, IL, USA) for analysis. Descriptive statistics were used for the sample. An exploratory analysis employing PCA was undertaken (with varimax rotation) to identify any factors within the 4 sections of LExps (12 items), LEnv (7 items), communication (9 items), and evaluation (15 items). Varimax rotation was used because our interest was in identifying whether the items developed for one construct were distinct from the items that would load on another construct. A loading of < 0.40 was employed as the cut-off. Any items exhibiting poor fit were eliminated. Poor fit was defined as any item that had a low communality score (< 0.40), did not load on any factor within a given section (< 0.40), or showed cross-loading on 2 factors (i.e., > 0.40 on more than 1 factor). PCA was re-run until a final factor solution was identified for each section. Next, tests of internal consistency utilizing the Cronbach alpha were calculated for each of the factors indicated by the PCA. The Cronbach alpha output indicating alpha values for the scale if items were deleted was used to reduce the number of items within each identified scale. Items were only removed if their removal would increase the alpha value or not cause a substantial reduction in the alpha value.