Institute for Public Opinion Research
School of Journalism and Mass Communications
Florida International University		 IPOR home page
 
 

Methodology

When writing stories please include the following:

This survey was conducted by the Institute for Public Opinion Research and the Cuban Research Institute of Florida International University. 1,201 randomly selected Cuban-American respondents were polled in Miami-Dade County and 606 in Broward. Interviews were completed on March 16, 2004. The margin of error for the overall poll is plus or minus 2.4%. This is the seventh in the series of FIU/Cuba Polls conducted since 1991.

 

The 2004 FIU/Cuba Poll

The seventh FIU/Florida Poll was conducted by the Institute for Public Opinion Research (IPOR) at Florida International University (FIU). From January 30 to March 15, 2004, 1,807 residents of Cuban descent living in Broward and Miami-Dade Counties were interviewed about their opinion of U.S. policies toward Cuba and other issues. The survey was funded by IPOR, the Cuban Research Institute (FIU), the Center for Labor Research & Studies (FIU), the South Florida Sun-Sentinel, and NBC6 TV WTVJ.

Results on selected issues were released and discussed in stories by the Sun-Sentinel, and NBC6 TV WTVJ on March 21 and 22. The full results of the poll will be posted on the IPOR web site and discussed at a press conference at FIU on March 31, 2004.

Sample

The Miami-Dade County sample was generated from telephone exchanges using standard random-digit-dialing procedures that result in each residential phone having an approximately equal chance of being chosen for the sample. Interviewers asked each household reached if a person of Cuban-American descent lived there who was eighteen years of age or older (28.9% of households in Miami-Dade County qualify acording to the 2000 U.S. census). The Miami-Dade sample was stratified into four areas (each composed of telephone wire center areas) that usually have different response rates. Interviews were conducted in each of the four areas until a sample proportional to the 2000 census count of Cuban-American households in each of the four areas was obtained. 1,201 respondents were interviewed in Miami-Dade County.

The Broward County sample was more difficult to obtain since only 3.1% of households in Broward belong to Cuban-Americans. Screening a RDD sample for qualifying households (as was done in Miami-Dade County) would not have been possible given funding constraints. The decision was made to use a listed sample and prescreen for Hispanic names. Phone numbers were selected randomly from the sample and the screening was done to reach Cubans 18 and older. In addition to the sample prescreened for Hispanic names, a smaller sample screening with all names listed (Hispanic and non-Hispanic) was drawn for six census tracts with a relatively high (over 20%) Cuban population. This was done to validate the prescreening procedure. Samples were obtained from Survey Sampling, Inc,. and the Hispanic name prescreening was done by them.

Sample Weights

Proportionally more interviews (606) were done in Broward County than the 3.1% Cuban share of the household count. For results reported for both counties together, results are weighted to ensure valid overall estimates. The ratio of female to male respondents in the survey was 58.8% to 41.2%. Results were weighted by gender/county to have estimates reflect the 2000 census ratios (50.6%/49.4% male/female for Broward; 48.5%/51.5% male/female for Miami-Dade).

Questionnaire

Bilingual (Spanish/English) interviewers conducted the interviews from IPOR'S telephone lab at the Biscayne Bay Campus of Florida International University. The questionnaire was programmed in both languages using IPOR's Sawtooth Ci3 Wincati system. 1,294 interviews were conducted in Spanish and 513 in English.

The format of the questionnaire is designed to make repeated tracking questions as comparable as possible. These questions are asked at the beginning of the survey in the same order each time the poll is done. Questions added to the poll in recent surveys or new this year are at the end. The questions from the 2000 poll can be seen at: http://www.fiu.edu/orgs/ipor/cuba2000/freq.htm

Cooperation and Response Rates

1,806 interviews were completed. 95 Cuban households reached (after screening) did not wish to be interviewed. 1,914 calls reached a household which had no person willing to do the interview (with 2 refusal conversion callbacks in most cases). It was not possible to determine which of these 1,914 households had a Cuban American resident. If we were to assume 28.9% of them were Cuban American (the Miami-Dade proportion) we would have a cooperation rate of approximately 0.52.

6,112 calls reached household answering machines on phones that were not picked up in 10 callbacks. With that number, we have an AAPOR response rate #3 of 0.183. Had the number of answering machines not picked up been 2,000 (closer to our expectation), we would have had a response rate #3 of 0.314. Unfortunately telephone system equipment problems resulted in FIU not being identified on caller id's of the answering machine telephones.

Principal Investigators

Dr. Guillermo J. Grenier is Professor of Sociology and Anthropology at Florida International University. Dr. Hugh Gladwin is Director of the Institute for Public Opinion Research (IPOR) at FIU and Associate Professor of Sociology and Anthropology.