TWO FACTOR ANOVA SAMPLE

1. In a previous related question, you undertook an analysis in which results showed that Development Practitioners who utilize Excel-spreadsheet during most of their business operations had lower number of errors in data entry during SPSS Trainings compared to Development practitioners who do not use the Excel in their business operations. A researcher would like to know if the same result extends to Development Practitioners in lower carders. The researcher planned a two-factor study comparing Excel Spreadsheet Users with non-users for Undergraduate, Master and PhD level qualifications consultants. For consistency across groups, grades were converted into six categories, numbered 0 to 5 from low to high. The results are presented in the following matrix.

1. How many study participants are in this study? n=24
2. How many variables are in this research problem? Three
3. What is the Dependent variable in this research problem? Number of errors in SPSS data
4. What is the Independent variable in this research problem? Excel usage and Consultants’ academic qualification
5. What is the measurement scale for the DV and IV

DV is measured on an interval scale while the IVs are measured on an ordinal scale.

• State three hypotheses being tested in this research problem?

H₀₁: Excel usage has no statistically significant effect on the number of errors in SPSS data

H₀₂: Consultant’s academic qualification level has no statistically significant effect on the number of errors in SPSS data.

H₀₃: The interaction between Excel usage and consultant’s academic qualification level has no statistically significant effect on the number of errors in SPSS data.

• Describe the appropriate statistical procedure for testing these hypotheses?

The appropriate procedure is the Two-Factor ANOVA

• Run the data using SPSS, and interpret the output

Descriptive statistics show that among undergraduate consultant, the mean number of error for non-users was 4 against 3 for users. In the master category, the mean number of error was 4 for non-users and 2 for users. In the PhD category, non-users had a mean number of errors of 4 while users had a mean of 1.

The main effect analysis indicate that academic qualification does not have a statistically significant effects on number of errors in SPSS data (F [2, 21] = 1.2, p= .324). On the other hand, Excel usage has a statistically significant effect on number of errors in SPSS data (F [1, 21] = 14.4, p<.01). The interaction effect analysis shows that the interaction between academic qualification and Excel usage does not have a statistically significant effect on the number of errors in SPSS data (F [1, 21] = 1.2, p= .324).

The results imply that when considered individually, Excel usage has a significant effect on SPSS data entry accuracy. However, the consultants’ academic qualification does have a statistically significant effect on SPSS data entry accuracy when considered on its own as well as when it is considered jointly with Excel usage.

1. Interpret the Levene’s test for equality of variance

The test shows that there is no statistically significant difference in the variances of the mean, median, median with adjusted df, and trimmed mean of the six groups of Excel users.

• Use F-table to verify results obtained using SPSS

The Table shows that the critical F-value for analysis of the main effect of academic qualification df (2, 21) is 3.47, which is higher than the calculated F value of 1.2. This supports the SPSS findings that academic qualification does not have a significant effect on number of SPSS errors.

The Table also shows that the critical F-value for analysis of the main effect of Excel usage df (1, 22) is 4.30, which is lower than the calculated F value of 14.4. This supports the SPSS findings that Excel usage has a significant effect on number of SPSS errors.

The Table also shows that the critical F-value for analysis of the interaction effect of academic qualification and Excel usage df (2, 21) is 3.47, which is higher than the calculated F value of 1.2. This supports the SPSS findings that the interaction between academic qualification and Excel usage does not have a significant effect on number of SPSS errors.

• From the results obtained, are the main effects for level of schooling statistically significant?

No

• From the results obtained, are the main effects of level of Excel Spreadsheet Use during business operations statistically significant?

Yes

• From the results obtained, are the interaction effects statistically significant?

No

• Prepare and submit write-up reporting the results using APA style according to the format attached in the appendix here.

Descriptive statistics show that among undergraduate consultant, the mean number of error for non-users was 4 against 3 for users. In the master category, the mean number of error was 4 for non-users and 2 for users. In the PhD category, non-users had a mean number of errors of 4 while users had a mean of 1. The main effect analysis indicate that academic qualification does not have a statistically significant effects on number of errors in SPSS data (F [2, 21] = 1.2, p= .324). On the other hand, Excel usage has a statistically significant effect on number of errors in SPSS data (F [1, 21] = 14.4, p<.01). The interaction effect analysis shows that the interaction between academic qualification and Excel usage does not have a statistically significant effect on the number of errors in SPSS data (F [1, 21] = 1.2, p= .324). The results imply that when considered individually, Excel usage has a significant effect on SPSS data entry accuracy. However, the consultants’ academic qualification does have a statistically significant effect on SPSS data entry accuracy when considered on its own as well as when it is considered jointly with Excel usage.