Chapter 5: Results
- 5 Results
- 5.1 Validity
- 5.2 Reliability
- 5.3 Regression Analysis
5.2 Reliability
The web-survey items are related to a construct. By means of the Cronbach’s alpha I will research whether the internal consistency of these items are sufficient for further analysis. For every construct its reliability will be analyzed.
Cronbach's alpha minimal requirement is 0,6 for further reliable analysis (Cooper et al., 2003). An alpha is mostly considered good when it is higher then 0,7. When it is possible to significantly improve the alpha by removing an item, often when the alpha improves by 0,05 or more, then I have to remove the item from the survey. In Table 5.1 are the results of the reliability test. No items are deleted.
| Construct | No. items | Cronbach’s alpha | N | |
|---|---|---|---|---|
| eWOM | 4 | 0,80 | 110 | |
| Information Sources | 7 | 0,83 | 110 | |
| Motivation | 3 | 0,89 | 110 | |
| Opportunity | 3 | 0,68 | 110 | |
| Ability | 3 | 0,87 | 110 | |
| Receiver's Purchase Intentions | 8 | 0,73 | 110 |
All missing values were deleted using listwise deletion (Hair et al., 1998) before I ran any of the tests. Listwise deletion involves removing incomplete cases (record with missing data on any variable) from the dataset. This means I had to remove all the records that had missing data on any variable. This resulted in a substantial decrease in the sample size available for the analysis (I went from 141 to 110 results), it does have important advantages. In particular, under the assumption that data are missing at random, it leads to unbiased parameter estimates and all analyses are calculated with the same set of cases.