I recently ran some data for my HLM class looking at student newspaper availability and political enthusiasm and participation. I had null findings everywhere. Here are more specifics:
Hypotheses
H1: The availability of a student newspaper on a college campus will positively impact individual-level enthusiasm towards the 2008 presidential campaign.
H2: The availability of a student newspaper on a college campus will positively impact an individual’s likelihood to attend a campaign event.
H3: The availability of a student newspaper on a college campus will positively impact an individual’s likelihood to volunteer for a campaign.
Sample estimates
Level 1: Survey of 25,000 college students at 50 colleges and universities in 4 battleground states during October 2008.
Level 2: Newspaper data, enrollment, public/private, and aggregate candidate contact data derived for 32 of the above schools.
Controls
Level 1: Sex, Age, Race, Candidate contact, Political ideology, Political party, news media use, and political discussion
Level 2: Enrollment, Public vs. private, and aggregate campaign contact.
IV
“Student newspaper availability” was created by multiplying the average print run for the paper by the number of times is pass printed per week and then dividing by undergraduate enrollment. This gave a value which was equivalent to the number of newspapers printed per undergrad per week.
DV’s
Enthusiasm: 4 point scale (ran as ordinal logit model)
Attend campaign event: Dichotomous
Volunteer for a campaign: Dichotomous
Findings
In none of the models did student newspaper availability have a significant effect on the outcome variable (p=.165 to p=.865).
Uncategorized
2008 election, HLM, journalism, newspaper, null, presidential, research
