| Option | Choice | Votes |
|---|---|---|
| Jacket | Tweed | 8 |
| Palette | Winter (Cool + Dark) | 9 |
| Pant | Joggers | 7 |
| Patterns | Keep it simple | 9 |
| Shoe | Crocs | 8 |
| Tie | Bow tie | 11 |
| Top | Sports jersey | 14 |
POLS 1140
Class Survey
Updated Mar 9, 2025
Wednesday
Plan
Today:
Sit with your groups
Overview of planned analyses
Take the draft survey
Propose any edits to the survey as needed
Friday:
- Draft planned analyses
Next Monday:
- Submit planned analyses
- Optional class
Friday
Plan
Take revised survey
Draft planned analyses
Upload to canvas by Monday
Zoom office hours next week
Attendance Survey!
Click here to have all of your most pressing questions answered.
Help me with my fit
Our Fashion Advice
What would Derek Guy say?
Apparently, khakis are in right now, but the pairing of tuxedo, crocs, bow tie and soccer jersey remains a timeless combination. Thanks https://t.co/BwjiEndcvH https://t.co/4ZHF1FlcMS pic.twitter.com/PmipDRJ3MB
— Paul Testa (@ProfPaulTesta) March 19, 2024
Planned Analyses
Overview
By Monday, please submit a document on Canvas covering the following:
Project summary
Research design
Formal expectations
Empirical tests
Project summary
A three sentence summary of your project:
We are interested in…
To study this question we do…
We expect to find
Research design
A list of the key components of your design. This should include:
Overview of any random assignments.
The dependent variable(s)
- Provide the question wording and response values (e.g. 1-100, 5 point scale, etc…)
- Are you using an individual item or a scale of items
The key predictor variable(s)
- If project contains an randomization, these will be treatment assignments
Other relevant predictors/covariates
Formal expectations
At least two formal expectations:
You can format these as hypotheses like H1 and H2 or as general expectations. However you write them, they should state in one sentence a clear, testable expectation.
Start with your most general expectation then move to more specific comparisons
Be specific. Instead of saying “We expect political knowledge to vary with media use” say, “H1: We expect that political knowledge will increase with news consumption”
Empirical tests
Translate these expectations into empirically testable implications.
Think in terms of means or averages, or regression coefficients.
- Maybe write down the formula of a regression model you’d like to estimate (feel free to use ChatGPT for help formatting)
If H1 is true, then the average level of political knowledge among high news consumption respondents should be higher than the average level of political knowledge among low news consumption respondents.
Revising the Survey
Overview
You can download the current pdf of the current draft of survey here
Then click here
Updates
Removed political participation measures
Combined Evaluations of Dem and Rep music preferences
Removed general questions about celebrity endorsements
Removed FT for journalists, academics and college students
Any specific edits to news articles/treatments
Group 1
I cut the learning question, if you want it back, think about another matrix question you could cut.
Randomized the order of perceived ideology and perceived trust. Do you think this will have an effect?
Group 2
Three treatment conditions:
- No information
- News article with Liberal source cue
- News article with Conservative source cue
How do you fill about the outcome measures?
Group 3
Two treatment conditions:
- Evaluate partisan music prefences first
- Evaluate you own music preferences first
If there’s an order effect, what would we expect?
Do you think we could combine the Democrat/Republican items into a single measure?
Group 4
- I think we may get to only ask one of the two long matrix questions
- which do you prefer?
- Any changes you would make?
- What variables from the rest of the survey do you think explain variation
Group 5
Two treatments:
- Celeb evaluations
- Celeb evaluations with stated endorsements
How do you expect evaluations to differ across these conditions?
Can we cut
g5_general?
Group 6
Any changes to these questions? Add
What predictors do you think will explain variation in response?
How should we code state of residence?
Group 8
Any additional responses/items you want to add to the matrix?
Do you envision a single scale (e.g. skepticism about the voter fraud) or individual items
What do you think will explain variation in people’s answers?
Group 9
Three treatment conditions:
- Neutral/Tax Foundation
- Liberal/NYT
- Conservative/WSJ
Any changes to the content of the treatment? Can we make it shorter? Punchier?
POLS 1140
POLS 1140 Class Survey Updated Mar 9, 2025