Perceptions and Plans for Prevention of Ebola: Results From a National Survey

Bridget Kelly; Linda Squiers; Carla Bann; Alexander Stine; Heather Hansen; Molly Lynch


BMC Public Health. 2015;15(1136) 

In This Article


Data Source

We recruited participants through an Internet panel maintained by GfK Custom Research, LLC. The GfK KnowledgePanel® consists of 50,000 adult panel members recruited by address-based sampling (ABS). The GfK KnowledgePanel® is based on probability sampling covering both online and offline populations in the U.S. GfK presents households with access to the Internet and a netbook computer, if needed. The resulting sample includes representation from listed and unlisted telephone numbers, telephone and non-telephone households, and cell phone-only households, as well as households with and without Internet access.

Sample Selection

Eligible participants were U.S. residents age 18 and older. A random sample of 3222 panel members was drawn from GfK's KnowledgePanel®. A total of 1018 participants completed the survey, yielding a final stage completion rate of 33 %. The panel recruitment rate for this study was 13.8 % and the profile rate was 64.1 %, for a cumulative response rate of 2.8 %. It is important to note that response rates for online panels tend to be lower than for other modes, due to the need to multiply recruitment rate, profile rate, cooperation rate and retention rate for a cumulative response rate.[16] We calculated response rates based on standard formulas for online panel response rates.[16]

Data Collection

The survey was completed as part of a larger survey with multiple topics and was fielded December 5–7, 2014. All procedures were approved by RTI International's Institutional Review Board. GfK obtains online consent from all panelists at the time they are recruited to the panel and again before each individual survey.


Measures are briefly described in this section. Additional file 1: Table S1 provides the full list of measures.

Perceived Susceptibility and Perceived Severity

Three measures assessed perceived susceptibility and severity. We asked how likely it was that the Ebola outbreak in West Africa would spread to the U.S. and how likely it was that the respondent or their community would be affected by Ebola. The measure of severity asked if someone in the respondent's community were to contract Ebola, how likely they would be to die from the disease.

Perceived Threat

We asked respondents to rate the level of perceived threat for each item on a list of issues, including heart disease, the seasonal flu, a pandemic flu (bird flu, swine flu), and Ebola.


Knowledge was measured with three items capturing how Ebola is spread and how long it could it take for someone to get sick after being exposed.

Behavioral Intentions

We measured behavioral intentions with a list of questions about social distancing and other protective measures.

Attitudes and Confidence

We also measured attitudes toward a number of Ebola-related policies such as mandatory quarantine and travel bans. We assessed confidence in media, government, and healthcare systems.

Other variables included respondent gender, age, race/ethnicity, education level, income level, U.S. Census region, and parental status. These items were collected as part of a profile completed by panelists upon joining the GfK KnowledgePanel®. These data are updated annually.


We calculated the percentage of respondents endorsing each attitude and confidence item by demographic characteristics. Demographic differences in responses were tested using logistic regression models to compute odds ratios adjusted for gender, age, education, race/ethnicity, income, presence of children in the home, and region. We conducted paired t-tests to compare differences in perceived threat and knowledge between Ebola and other issues. The survey data were weighted to represent the U.S. population based on the most recent Census reports. The survey weights were developed using an iterative proportional fitting procedure, utilizing the following demographic characteristics: gender, age, income, race/ethnicity, region, metropolitan statistical area (MSA), and Internet access. The survey weights were then incorporated into our statistical analyses using the survey procedures in SAS version 9.3.