Pairing Big Data in Bite Size Morsels
•Download as PPTX, PDF•
0 likes•59 views
Presented this slide deck to analytic and evaluation professionals at the Ohio Program Evaluators' Group's bi-annual conference. Discussed how to reduce large, complex datasets into smaller, manageable projects.
Recommended
More Related Content
What's hot
What's hot (14)
Similar to Pairing Big Data in Bite Size Morsels
Similar to Pairing Big Data in Bite Size Morsels (20)
More from Michael Levin
More from Michael Levin (14)
Recently uploaded
Recently uploaded (20)
Pairing Big Data in Bite Size Morsels
- 1. Pairing Big Data in Bite Size Morsels Dr. Michael A. Levin Otterbein University @MichaelALevin
- 2. Pairing Big Data in Bite Size Morsels Nature of Data Descriptive Look at Your Data Correlation Sampling Putting Together
- 3. Data Appear Differently Nominal Ordinal 1. 2. 3.
- 4. Data Appear Differently Interval Ratio 1 2 3
- 5. Descriptive Statistics Data Permissible Statistics Nominal Description Mode Ordinal Order Median Interval Distance Mean, Std Dev, Skew, Kurt Ratio Zero Harmonic mean
- 6. Look at Your Data
- 8. Look at Your Data 1 2 3 4 5
- 9. Look at Your Data
- 10. Look at Your Data
- 14. Putting it All Together
- 15. Putting it All Together
- 16. Pairing Big Data in Bite Size Morsels Dr. Michael A. Levin Otterbein University @MichaelALevin
Editor's Notes
- Clean your data. All datasets need a good cleaning.
- Experts can help you reduce columns or variables. Also, they can help makes sense of the data or the underlying model.
- Look for lack of variance in your responses.
- Look for ties. Too many ties creates problems when running Spearman Rank Rho
- Handle missing and inconsistent data.
- Are you measuring the same variables or constructs or ideas twice?
- Treat your large dataset as a known population and calculate a mean and a standard deviation for a variable of interest. Create a confidence interval or mean based sample size.