28
Internet Usage Statistical Data Analysis Edgardo Donovan RES 610 – Dr. Joshua Shackman Module 5 – Session Long Project Monday, September 19, 2011

Internet Usage Statistical Data Analysis

Embed Size (px)

DESCRIPTION

Internet Usage Statistical Data Analysis

Citation preview

Page 1: Internet Usage Statistical Data Analysis

Internet UsageStatistical DataAnalysis

Edgardo DonovanRES 610 – Dr. Joshua ShackmanModule 5 – Session Long Project

Monday, September 19, 2011

Page 2: Internet Usage Statistical Data Analysis

Overview 1. Study Background 2. Top/Lowest Uses 3. Top/Lowest Uses Chart 4. InterSurvey 5. Sampling Issues 5. Sampling Issues 6. Hypotheses 7. Case Processing 8. Reliability 9. Item Statistics 10. Item Statistics (cont.) 11. Summary Item Stats

Page 3: Internet Usage Statistical Data Analysis

Overview (cont.) 12. Inter-Item Correlations 13. More Quantitative Analysis 14. Item Statistics 15. Item Statistics (cont.) 16. Improving the Original 16. Improving the Original 17. Improved USC Model 18. Autocorrelations 19. Hours on the Internet 20. Hours on Internet (cont.) 21. Positive Correlations 22. Positive Correlations (cont.)

Page 4: Internet Usage Statistical Data Analysis

Overview (cont.) 23. Conclusion 24. Questions?

Page 5: Internet Usage Statistical Data Analysis

1. Study Background

2000 UCLA study surveying the digital future

Limited to WebTV users Initially started at Stanford, then UCLA, Initially started at Stanford, then UCLA,

then USC

Page 6: Internet Usage Statistical Data Analysis

2. Top/Lowest Uses

Top Uses: Learning Surfing (overlap?) Reading about products Reading about products

Surprising Lowest Uses: Schoolwordk Banking Job Search

Page 7: Internet Usage Statistical Data Analysis

3. Top/Lowest Uses Chart

Page 8: Internet Usage Statistical Data Analysis

4. InterSurvey

Relied upon a form application tool named “Intersurvey”

Survey had to be done online Low interest in effectively sampling the US Low interest in effectively sampling the US

Internet user population

Page 9: Internet Usage Statistical Data Analysis

5. Sampling Issues

WebTV Set Top Boxes Limited to low end income demographic

Poor attempt at sampling External validity problematic External validity problematic

Page 10: Internet Usage Statistical Data Analysis

6. Hypotheses

“Negative correlation between Internet and TV use

Negative correlation between Internet Use and traditional social activity and shoppingand traditional social activity and shopping

No insight on survey questions

Page 11: Internet Usage Statistical Data Analysis

7. Case Processing

Case Processing Summary

N %

Cases ValidCases Valid1241 100.0

Excludeda

0 .0

Total1241 100.0

a. Listwise deletion based on all variables in the procedure.

Page 12: Internet Usage Statistical Data Analysis

8. Reliability

Reliability Statistics

Cronbach's AlphaCronbach's Alpha Based on

Standardized Items N of Items

.816 .809 17

Page 13: Internet Usage Statistical Data Analysis

9. Item Statistics

Page 14: Internet Usage Statistical Data Analysis

10. Item Statistics (cont.)

Page 15: Internet Usage Statistical Data Analysis

11. Summary Item Stats

Page 16: Internet Usage Statistical Data Analysis

12. Inter-Item Correlations

Page 17: Internet Usage Statistical Data Analysis

13. More Quantitative Analysis

Page 18: Internet Usage Statistical Data Analysis

14. Item Statistics

Page 19: Internet Usage Statistical Data Analysis

15. Item Statistics (cont.)

Page 20: Internet Usage Statistical Data Analysis

16. Improving the Original

USC to improve the Stanford/UCLA study Auto, Pharms, and groceries were removed Smoothing effect 10-25 questions that delve deeper into 10-25 questions that delve deeper into

issues

Page 21: Internet Usage Statistical Data Analysis

17. Improved USC Model

Page 22: Internet Usage Statistical Data Analysis

18. Autocorrelations

Page 23: Internet Usage Statistical Data Analysis

19. Hours on the Internet

Page 24: Internet Usage Statistical Data Analysis

20. Hours on Internet (cont.)

Page 25: Internet Usage Statistical Data Analysis

21. Positive Correlations

Positive correlation between hours spent on the Internet and amount of online purchases

Significant deviation between males and females concerning when purchasing Significant deviation between males and females concerning when purchasing sporting goods

Page 26: Internet Usage Statistical Data Analysis

22. Positive Correlations (cont.)

Page 27: Internet Usage Statistical Data Analysis

23. Conclusion

Stanford/UCLA Study WebTV and InterSurvey Limitations Extra Variables Measured “strange” usage Measured “strange” usage

USC Eliminated Unnecessary Variables No WebTV InterSurvey Limitations Hypothetical Correlations Have Value More advanced Stage of Internet Use

Page 28: Internet Usage Statistical Data Analysis

24. Questions?

Questions?

Edgardo Donovan Trident University [email protected]