How many registered voters would vote to elect a female president? A curious pol

ID: 3359430 • Letter: H

Question

How many registered voters would vote to elect a female president? A curious political scientist selected two equivalent samples of voters. One group was given the description of a hypothetical male candidate who takes the most popular stance on most major issues, while the other group was given this same description while being told that the candidate was female. The two groups were then asked to rate on a scale from 1 to 5 the likelihood that they would vote for their respective candidate (with 1 being least likely and 5 being most likely). The following scores were obtained: Told Candidate Was a Man Told Candidate was a Woman 4 4 Applying the median test, determine whether there is a significant difference between the medians of these groups. Was the likeihood of this hypothetical candidate being voted for influenced by his or her gender?

Explanation / Answer

Result:

MINITAB used.

Mood's Median Test: score versus group

Descriptive Statistics

group

Median

N <= Overall Median

N > Overall Median

Q3 – Q1

95%
Median
CI

4.0

(3, 5)

2.5

(2, 3)

Overall

3.0

95.0% CI for median(1) - median(2): (0,3)

Test

Null hypothesis

H: The population medians are all equal

Alternative hypothesis

H: The population medians are not all equal

Chi-Square

P-Value

6.17

0.013

Calculated chi square =6.17 which is > 3.841 .

Ho is rejected

There is a significant difference between the two groups.

We conclude that Voting is influenced by gender.