Assignments and Assessment
Assignments are stages of one professional survey workflow. Students will demonstrate individual methodological understanding, individual ability to analyze survey data, and effective collaboration on a public-facing research product. Detailed prompts, rubrics, templates, and submission links will be released on Canvas.
Assessment map
| Category | Weight | Responsibility |
|---|---|---|
| Preparation and professional participation | 10% | Individual |
| R and statistics labs | 25% | Individual, following guided class work |
| Individual methods assignments | 15% | Individual |
| Univision project milestones | 15% | Team |
| Individual poll analysis memo | 10% | Individual |
| Coauthored report and presentation | 20% | Class and section teams |
| Reflection and contribution statement | 5% | Individual |
| Total | 100% |
R and statistics labs · 25%
Five substantial labs integrate R programming, statistical reasoning, and survey analysis:
- R, Quarto, importing, cleaning, recoding, and documenting survey data.
- Descriptive statistics, cross-tabulations, measurement, and visualization.
- Sampling, simulation, margins of error, confidence intervals, and basic hypothesis tests.
- Survey weights, design effects, and weighted estimation.
- Regression, predicted probabilities, graphics, and reproducible reporting.
Most lab work will be completed together during class. Each lab concludes with a shorter On Your Own section completed alone. Labs are graded on statistical reasoning, readable and functioning code, reproducibility, appropriate checks, and clear interpretation.
Initial Survey Autopsy diagnostic
During the first class on August 24, all students will evaluate the same intentionally problematic survey before receiving formal instruction in survey methodology. This short exercise is completion-based and is used to identify the class’s starting knowledge. It is not Individual Assignment 1.
Open the Initial Survey Autopsy instructions and download the report
Assignment materials
Individual methods assignments · 15%
Three concise assignments ask students to make and defend methodological decisions.
1. Poll autopsy · 5%
Assigned: October 12 · Due: October 18 at 11:59 p.m.
Evaluate a published poll from its public release and methodological disclosure. Identify the target population, sample source, mode, field dates, question wording, weighting, reported uncertainty, principal claims, and important information that is absent. Conclude with what the poll does and does not permit readers to claim.
2. Measurement and questionnaire critique · 5%
Diagnose a set of survey questions using principles of comprehension, retrieval, judgment, response mapping, question order, response-option design, and sensitive measurement. Revise selected questions and explain how the revisions improve validity, reliability, comparability, or respondent experience.
3. Sampling and mode recommendation · 5%
Given a research objective, population, budget, timeline, and reporting needs, recommend a sample and data-collection design. Address coverage, recruitment, nonresponse, language access, feasibility, weighting, uncertainty, and the limitations that should accompany published results.
Univision project milestones · 15%
Milestones make the collaborative process visible and keep the live project on schedule. They are working products rather than unrelated homework.
| Milestone | Collaborative deliverable |
|---|---|
| Project charter | Objectives, audiences, claims, roles, decision rules, confidentiality, and timeline |
| Measurement package | Construct map, proposed question module, source questions, and design rationale |
| Testing package | Cognitive-testing protocol, findings matrix, English/Spanish revisions, and unresolved issues |
| Representation plan | Target population, sampling and mode assessment, recruitment, language access, and limitations |
| Instrument QA | Integrated questionnaire review, test cases, mobile/bilingual checks, issue log, and approval record |
| Analysis plan | Outcomes, subgroups, derived variables, weighting use, planned tables and figures, and decision rules |
Milestones will normally be evaluated with a short common rubric:
- methodological justification;
- completeness and accuracy;
- use of evidence and testing;
- documentation and version control;
- responsiveness to feedback; and
- readiness for the next production stage.
Individual poll analysis memo · 10%
Each student will independently answer one focused substantive question using the final poll data.
The memo will normally include:
- a clearly stated research question;
- a brief explanation of the relevant survey measures;
- documented cleaning and recoding decisions;
- appropriate use of survey weights;
- descriptive estimates and uncertainty;
- one justified subgroup comparison or introductory model;
- at least one publication-quality visualization;
- a concise interpretation for a policy or public audience;
- limitations; and
- reproducible Quarto source code.
Students may work from a shared, quality-controlled analysis file, but the analytical choices, code, verification, and prose in the memo must be their own.
Individual reflection and contribution statement · 5%
The final reflection will:
- document the student’s specific contributions;
- identify evidence of those contributions in the project record;
- explain one consequential methodological decision;
- evaluate the strengths and weaknesses of the completed project;
- describe what the student learned about collaboration; and
- propose one improvement for a future polling cycle.
The contribution statement provides accountability, but it is not a competition over who reports the greatest number of tasks. Quality, reliability, documentation, and support for the team matter.
Evaluation principles
Across assignments, work will be evaluated on:
- methodological reasoning;
- statistical accuracy;
- reproducibility and documentation;
- clarity and audience awareness;
- ethical and professional judgment;
- responsiveness to evidence and feedback; and
- contribution to the collaborative process.
A polished output cannot compensate for an indefensible method, and technical sophistication cannot compensate for unclear communication.