I provide online causal inference and program evaluation tutoring for graduate students in the Los Angeles and San Francisco Bay Area metros. I regularly work with students from UCLA, USC, UC Irvine, and Caltech, as well as UC Berkeley, Stanford University, UC San Francisco (UCSF), and other UC and private graduate programs. All tutoring is delivered online; I do not maintain a physical office in these cities.
Causal inference and program evaluation are core components of graduate training in economics, public policy, public health, education, sociology, and applied statistics. Students often struggle not with running regressions, but with understanding identification: when a coefficient can be interpreted causally, what assumptions are required, and how to communicate those assumptions clearly without overstating results.
I help you move from “I ran the model” to “I can defend the design.” That includes choosing the right identification strategy, understanding threats to validity, interpreting estimates correctly, and writing a graduate-level methods section that holds up under scrutiny.
Speak Directly With the Tutor
If you’re unsure whether your results are causal, confused about identification assumptions, or stuck choosing between designs, reach out directly. You’ll speak with the tutor who actually works through the analysis with you.
Call/Text: 510-398-0006
Email: tutor@californiagraduatetutor.com
What Causal Inference & Program Evaluation Tutoring Covers
- Potential outcomes framework and causal estimands
- Randomized experiments and internal validity
- Difference-in-differences (DiD)
- Regression discontinuity designs (RDD)
- Instrumental variables (IV)
- Matching and selection on observables
- Threats to identification and robustness checks
- Interpreting treatment effects responsibly
- Writing defensible methods and results sections
Core Causal Framework (MathJax Standard)
In the potential outcomes framework, each unit \(i\) has two potential outcomes:
\[ Y_i(1), \; Y_i(0) \]
The individual causal effect is:
\[ \tau_i = Y_i(1) – Y_i(0) \]
Since we never observe both outcomes for the same unit, causal inference relies on assumptions that allow us to estimate average treatment effects.
Difference-in-Differences (Conceptual Form)
A standard two-group, two-period DiD estimand is:
\[ \text{DiD} = (\bar{Y}_{T,1} – \bar{Y}_{T,0}) – (\bar{Y}_{C,1} – \bar{Y}_{C,0}) \]
We focus heavily on the parallel trends assumption, how to assess its plausibility, and how to discuss limitations clearly in graduate work.
A Graduate-Level Program Evaluation Workflow
- Define the policy or treatment clearly
- Specify the causal estimand of interest
- Choose an identification strategy
- State assumptions explicitly
- Estimate effects carefully
- Test robustness and sensitivity
- Interpret results without overclaiming
- Write a methods section that can be defended
Common Graduate Use Cases
- Policy evaluation papers in economics or public policy
- Applied causal analysis in public health or education
- Replication of published program evaluation studies
- Thesis or dissertation methods chapters
- Course projects requiring quasi-experimental designs
Software Support
- R: causal workflows and interpretation
- Stata: DiD, IV, RDD, and post-estimation analysis
- SPSS: regression-based program evaluation
Related Statistics & Research Support
Need Help Making a Causal Claim You Can Defend?
If you’re unsure whether your design supports a causal interpretation, I can help you clarify assumptions, strengthen your analysis, and explain results clearly at a graduate level.
Call/Text: 510-398-0006 | Email: tutor@californiagraduatetutor.com