SPSS Descriptive Statistics & Reliability

In research paper you can show descriptive statistics of your demographic data and descriptive statistics of your scale items. In this example, in your assignment we will only show descriptive statistics for scale items. However, in your research paper you should also show your demographic data before starting your data analysis part.

Example for demographic data (This table independent from the data file CAMdata01.xls):

Survey responses were captured using a 7-point Likert scale, where 1 = Strongly Disagree, and 7

= Strongly Agree. The sample sizes were N = 180 (Pre-Test survey) and N = 159 (Post-Test survey). The mean responses and the standard deviations (in parentheses) for each item are shown below.

Provide the descriptive statistics of each scale item in Table 2 by using data file CAMdata01.xls as described below:

Tip: In Table 2, descriptive statistics for “Ease of Statistics (EA)” provided as an example. Also, you can check correction of your analysis from this example.

Using SPSS:

1.1. Start IBM SPSS Statistics 21/22. Select File > Open > Data > files of type=Excel. Open CAMdata01.xls. Select Analyze > Descriptive Statistics > Descriptives. Select the 11 “Pre” variables. Click OK.

1.2. Repeat the same steps in the previous part (1.1) for “Post” variables.

2

4

2. Reliability (15 points)

OVERVIEW: IN THIS PART YOU NEED TO FILL UP TABLE 3 AND YOU NEEDD TO PROVIDE RELAIBILITY ANALYSIS FOR OTHER VARIABLES LIKE IN 2.1 EXAMPLE.

Information: Internal consistency across different survey items (scale items) attempting to measure the same construct is typically measured using Cronbach’s alpha, when the responses are on a Likert scale. For categorization studies using multiple raters use Cohen’s kappa (for two raters) or Fleiss’ kappa (for more than two raters).

These values are assessed against the following commonly accepted guidelines (George and Mallery 2003, p. 231): Excellent in [1, 0.9], Good in (0.9, 0.8], Acceptable in (0.8, 0.7], Questionable in (0.7, 0.6], Poor in (0.6, 0.5], and Unacceptable in the (0.5, 0] range of alpha values.

Pre-test constructs (N = 180) as well as post-test constructs (N = 159) are listed below. Pre-survey and post survey items were identical, with the exception of a fourth usefulness item (US4) added to the post-survey, which was a repetition of US2.

Table 3. Reliability Analysis

Reliability

Construct Survey Items Cronbach's Alpha Alpha Assessment

Provide the reliability statistics using data file CAMdata01.xls as described below:

Tip: Reliability analysis of “Ease of Statistics (EA)” in Pre-Survey is provided as an example.

Using SPSS:

2.1. Select Analyze > Scale > Reliability Analysis. Select a group of variables corresponding to the same construct, e.g. PreEA1-PreEA3. Click Statistics. Check Descriptives for = Scale if item deleted. Select Continue > OK.

Reliability Statistics

Item-Total Statistics

The results show that the reliability using all three items is equal to 0.903. Deleting any of the three items would result in reliability reduction. Therefore, we should keep all three.

2.2. Repeat the same steps in the previous part (2.1) for each of the variables in Table 3. And fill up the Cronbach's Alpha and alpha assessment values in Table 3. Also, provide your finding for each construct reliability statistics as shown above example.

Tip: Do reliability analysis only for pre-survey items.

Tip: Show your “Reliability Statistics” for each variable like in 2.1.

Tip: Fill up the Cronbach's Alpha and alpha assessment values in Table 3.

References

George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and reference.

11.0 update (4th Ed.) Allyn & Bacon, Boston.