public-opinion-media-analysis-1
Survey Assignment 3: Framing and Public Opinion
Due November 13 at 1pm (Canvas)
Purpose: To explore the relationship between media coverage and public opinion of/attention to impeachment. This is using data that YOU created in class, so the assignment is the logical conclusion of the data collection we did during the last week in October.
Instructions
Use the data here (make sure to copy it to your Google Drive) and the codebook below to analyze data on media coverage of impeachment.
- Analyze the average amount of coverage by all sources, liberal sources, and conservative sources. Using relevant materials covered in class (cite your sources), discuss why the patterns (if any) of coverage between sources likely look the way they do.
You can most easily create the table below by doing a pivot table, but not selecting any row or column variables. Just put in the average values for each of the sources (“3day”). E.g.:
A table presenting the results should look something like the following (feel free to rip this table formatting, but not the numbers, off):
Source | Average Coverage (These numbers are made up) |
All | 4,815 |
Conservative Outlets | 162 |
Liberal Outlets | 342 |
- Pick a frame type (politics, info, events, etc) and create a table similar to the one in #1 detailing that frame type’s prevalence across sources. Interpret the findings as per #1.
- Analyze the correlation between some public opinion variable (% of individuals supporting impeachment, % of Republicans opposed, etc) and the amount of coverage of impeachment (across all sources, liberal sources, etc). Do this twice (e.g., two pairs of variables) and explain what you found, using material from the class to help explain what you found.
This is new to you. Quick explainer: a correlation value is a numeric summary of the relationship between two variables, and ranges from -1.0 to +1.0. Values closer to -1 and +1 indicate a stronger negative/positive relationship and values at or near 0 indicate no correlation.
A positive correlation indicates that as one variable increases in value, the other tends to as well. E.g., as people become more educated, they tend to make more money.
Negative correlations indicate that increases in one variable are associated with decreases in the other. E.g., as income levels increase, crime rates tend to decrease.
To do a correlation, select two columns of data using the =CORREL function:
In this case, I have selected all of the rows for liberal coverage of events (column X, rows 2:37) and the % of Republicans who support impeachment (column C, rows 2:37). It doesn’t matter which order you enter the two variables.
After hitting enter I get a correlation of -.11, which is a slight negative relationship between the two.
- Finally, conduct an analysis of public attention to impeachment (measured using Google search trends) and one of the media variables in the dataset. Do this using the same correlation test you used in #3, and explain your findings, being sure to use materials in the class to interpret what you found.
Guidelines
The paper should be around 3 pages, inclusive of tables, in double-spaced 12-point font
Each of the analyses (1-4 above) is worth 20% each
The use of class materials/citations is worth 20%
Data Codebook
Variable Name | Description |
yes | % of individuals saying Trump should be impeached |
rep_yes | % of Republicans saying Trump should be impeached |
dem_yes | % of Democrats … |
ind_yes | % of independents … |
no | % of individuals saying Trump should not be impeached |
rep_no | % of Republicans … |
dem_no | % of Democrats … |
ind_no | % of independents … |
all_sources_3day | Average (3-day) amount coverage of impeachment, all sources |
lib_3day | Average amount of impeachment coverage, NYT and MSNBC |
con_3day | Average amount of impeachment coverage, Fox and NYPost |
cumulative_coverage | Cumulative amount of (all) media coverage |
frames_info | Number of informational frames about impeachment |
frames_politics | Number of political (e.g., political fighting, election-related discussion, etc) frames about impeachment |
frames_events | Number of event-related frames (whistleblowers, revelations coming to light, etc) |
frames_comm | Communication frames (Trump Twitter coverage, speeches, etc) |
frames_legal | Legal frames (is quid pro quo a crime, legal grounds for impeachment, etc) |
frames_*_con | [Frame type] from conservative media |
frames_*_lib | [Frame type] from liberal media |
google_searches | Index (0-100) of Google searches on impeachment |