| Age: M = 44.3, SD = 10.1 | |||
| Years of Experience: M = 18.4, SD = 9.4 | |||
| Industry | Job | Race/Ethnicity | N (%) |
|---|---|---|---|
| Tech / Ads | Senior Data Scientist | Asian | 1 (14.3) |
| University | Associate Professor | Asian, I do not identify with a particular race or ethnicity | 1 (14.3) |
| Healthcare | Senior machine learning engineer | Latino/a, Latinx, or Latine, Hispanic, White | 1 (14.3) |
| Academic | Research Scientist | White | 4 (57.1) |
| Education | Professor | ||
| Higher Ed | Postdoctoral Reseacher | ||
| product development/ux | principle uxr | ||
What Is Analytic Fluency?
Matthew E. Vanaman
What Analytic Fluency Might Be
- Our provisional definition:
- The “soft skills” of data analysis that go beyond formal training in statistics and which are thought to come from experience analyzing data “in the real world”.
The Present Research
Goals
- Make analytic fluency explicit.
- Check our intuitions.
- Develop a theory of analytic fluency.
Goals
- Long term: develop a grounded theory of analytic fluency
- Short term: pilot qualitative study
Method
Interview Techniques
- Focus groups 2x
- One-on-one semi-structured interviews 14x
- Semi-structured questionnaire
Method
Example Questions
Imagine you had to teach a course called “data analysis in the real world”. What would be the major themes or take-home lessons you would want your students or trainees to leave with?
Reflect back to when you were first starting out. Now that you have had all your years of experience analyzing data in the wild, what advice would you give your less-experienced self?
Method
Focus Group Sample
Method
Analysis
Thematic content analysis
- Round 1: Open coding
[I think] someone straight out of college might know how to do perfect science, but kind of [go] overboard with methodology.
- Open code: “doesn’t overcomplicate”
Method
Analysis
Thematic content analysis, round 2: Identify themes from open codes
Example open codes:
- tailors analysis to stakeholder
- reporting must be relevant to stakeholder
- use narrative to explain results/data
- stakeholder impact trumps rigor
Theme: “Prioritizes Stakeholder Needs”
Results
Provisional Themes
Results
Provisional Themes
Future (Short-Term)
1-on-1 Semi-Structured Interview Demographics
| Age: M = 45.8, SD = 10.3 | |||
| Years of Experience: M = 18.5, SD = 9.1 | |||
| Industry | Job | Race/Ethnicity | N (%) |
|---|---|---|---|
| Technology (social media specifically) | Quantitative UX Researcher | Asian | 1 (16.7) |
| Academia | Professor and Program Director | White | 5 (83.3) |
| Academia | Professor | ||
| Academia | Associate Professor | ||
| Academia | Associate Professor | ||
| Higher Ed | Research Analyst | ||
Future (Short-Term)
Summer goals:
- more rigorous thematic content analysis
- analyze both focus group and SSI data
Future (Long-Term)
- More systematic grounded theory approach
- Self-report measure of analytic fluency
- Identify antecedents and consequences of analytic fluency
Future (Long-Term)
Computer adaptive testing is so cool
Thank you!
Questions?