Reducing Teen Births among Latinas: A Pilot RCT to Test the Effectiveness of Radio Messages

The central hypothesis is that a radio message from a Hispanic teen that stresses the opportunity costs to the mother of having a baby (e.g., missed opportunities for fun, socializing with friends, or attending school) will diminish births to Hispanic teens.

Our sample consists of the coverage areas of 28 FM radio stations (mostly Spanish-language) in California. We randomly assigned 14 stations to the treatment group and asked them to air a 30-second ad 15 times per week for five weeks. Thirteen of the 14 stations agreed to air the ad. On most of these stations, the ad aired during a five-week period with one of three possible start dates (August 31, Sept. 14, or Sept. 28). However, one station began and ended early (airing the ad from August 4 to Sept. 6), and another station missed one spot and made up for it on Nov. 14. Thus, the earliest date that the ad was aired on any station was August 4, 2015, and the latest date was Nov. 14, 2015.

2015-08-04

2015-11-14

For each of the 28 radio station coverage areas, our primary outcome is the number of births from Jan. 19, 2016 (5.5 months after the first airing of the ad) to Sept. 14, 2016 (10 months after the last airing) to Hispanic females age 20 or younger, divided by the number of births from Jan. 18, 2015 to Sept. 14, 2015 to Hispanic females age 20 or younger.

The coverage areas of 28 FM radio stations (mostly Spanish-language) in California are included in the study. We randomly assigned 14 of the 28 radio stations to the treatment group and asked each treatment group radio station to air a 30-second ad 15 times per week for five weeks. In the ad, a young woman stressed the opportunity costs to teen mothers of having a baby (e.g., missed opportunities for fun, socializing with friends, or attending school). The study will estimate the effects of the intervention on births to female Hispanic teenagers in the year after the ad was aired, using birth records from the California Department of Public Health.

Randomization done in office by a computer

The unit of randomization is the radio station. (Since the data for estimation of treatment effects will consist of one observation for each radio station's coverage area, we have not classified the treatment as clustered in our answer to the next question.)

No

28 radio station coverage areas

28 radio station coverage areas

14 radio station coverage areas assigned to treatment, 14 radio station coverage areas assigned to control

Pre-Analysis Plan

This study uses a randomized controlled trial (RCT) to evaluate the effectiveness of a mass media campaign designed to reduce pregnancy among Latina teenagers. We aired public service announcements (PSAs) on terrestrial radio stations in California. The sample includes 28 radio stations and over 11,000 births to Latina teenagers in 2016. The PSA, narrated by a young Latina, stressed the opportunity costs to teenage mothers of having a baby, e.g., missed opportunities for fun, socializing with friends, or attending school. Although this message expresses themes articulated by Latina teens and young mothers whom we interviewed, geocoded birth records suggest the PSAs had little effect. With data on all births in the state obtained from the California Department of Public Health, we can reject the hypothesis that the PSAs reduced births to Latina teenagers by 9% or more based on the 90% confidence interval of the estimated average treatment effect. In addition to shedding light on this particular radio campaign, this study provides a template for the design and analysis of mass media campaigns to improve other public health outcomes.

Green, D.P., Zelizer, A. & Lin, W. How Effective Are Radio Messages Aimed at Reducing Teen Births Among Latinas? A Randomized Controlled Trial. Popul Res Policy Rev (2020). https://doi.org/10.1007/s11113-020-09583-x

https://rdcu.be/b36NC