Using the Campaign Media Dataset, we are interested in understanding what causes variation in the tone of articles that mention Justin Trudeau. One hypothesis is that the more often Trudeau is mentioned, the more positive the tone of the article. The relationship between article tone and the frequency with which Trudeau is mentioned is presented in Table 1. Below the table are summary statistics. Review both Table 1 and the summary statistics and respond to the questions that follow.
Table 1. Article Tone by Frequency of Trudeau Mentions in the Article
Frequency of Trudeau Mentions in the Article | |||
Article Tone | Low | Moderate | High |
Negative | 56% | 63% | 49% |
Neutral | 23 | 21 | 21 |
Positive | 21 | 16 | 30 |
Total | 100% | 100% | 100% |
(N) | (71) | (67) | (53) |
Summary Statistics for Table 1
Measure of Association | Value | Statistical Significance (p-value) |
Phi | 0.14 | 0.459 |
Cramer’s V | 0.10 | 0.459 |
Gamma | 0.09 | 0.419 |
Tau b | 0.05 | 0.419 |
Tau c | 0.05 | 0.419 |
1. Describe the relationship in Table 1 in a paragraph (5-7 carefully-worded sentences). Do not refer to measures of association.
2. Which measure of association is appropriate for these data, and why? (2-3 carefully-worded sentences)
3. Based on the evidence in Table 1 and the summary statistics, is there any support for the hypothesis that the more often Trudeau is mentioned, the more positive the tone of the article? Be sure to interpret and make reference to the appropriate measure of association and statistical significance. (4-5 carefully worded sentences)
Section 2
Another possibility is that the source of the articles matter. One hypothesis is that the Toronto Star generally has a more positive tone than the Globe and Mail regarding Justin Trudeau. The relationship between article tone and source is presented in Table 2. Below the table are summary statistics. Review both Table 2 and the summary statistics and respond to the questions that follow.
Table 2. Article Tone by Source
Source | ||
Article Tone | Globe and Mail | Toronto Star |
Negative | 59% | 55% |
Neutral | 23 | 21 |
Positive | 18 | 24 |
Total | 100% | 100% |
N | (80) | (111) |
Summary Statistics for Table 2
Measure of Association | Value | Statistical Significance (p-value) |
Phi | 0.07 | 0.656 |
Cramer’s V | 0.07 | 0.656 |
Gamma | 0.09 | 0.473 |
Tau b | 0.05 | 0.473 |
Tau c | 0.05 | 0.473 |
4. Describe the relationship in Table 2 in a paragraph (5-7 carefully-worded sentences). Do not refer to measures of association.
5. Based on the evidence in Table 2 and the summary statistics, is there any support for the hypothesis that articles in the Toronto Star have a more positive tone than articles in the Globe and Mail? Be sure to interpret and make reference to the appropriate measure of association and statistical significance. (4-5 carefully worded sentences)
6. Which relationship is stronger: the one in Table 1, or the one in Table 2? Refer to the appropriate measure(s) of association in your answer. (1-2 carefully-worded sentences)
Section 3
Perhaps the relationship between how frequently Justin Trudeau is mentioned in the article and the tone depends on a third variable: the source. One hypothesis is that the more often Trudeau is mentioned, the more positive the tone of the article. To examine this, we look at the relationship between article tone and frequency of Trudeau mentions in Globe and Mail articles (Table 3A) and Toronto Star articles (Table 3B). Below each of these tables are the summary statistics. Review these and respond to the question that follows.
Table 3A. Article Tone by Frequency of Trudeau Mentions, Globe and Mail
Frequency of Trudeau Mentions in the Article | |||
Article Tone | Low | Moderate | High |
Negative | 53% | 56% | 74% |
Neutral | 26 | 22 | 16 |
Positive | 21 | 22 | 10 |
Total | 100% | 100% | 100% |
(N) | (34) | (27) | (19) |
Get Free Quote!
308 Experts Online