Archive - Survey of Older Minnesotans
2005 Survey of Older Minnesotans
The Survey of Older Minnesotans was collected via telephone interviews with a random sample of noninstitutionalized persons aged 50 and over in the state of Minnesota. The sample was stratified into Metro and Non-Metro regions; independent samples were drawn for each region so that the total number of completed interviews in each could be controlled, allowing for both analyses of both the regional and merged samples. The final merged sample size was 3,855 persons. The margin of error for this sample is less than 2%.
Interviews were conducted between May 13 and June 20, 2005. The average interview length was 20.1 minutes. Sampled telephone numbers were called according to a schedule designed to minimize bias problems associated with difficult-to-reach respondents. Each number in the sample was called up to 12 times, or until the interview was completed or the respondent declined to be interviewed.
Survey data were entered and automatically consolidated into a CATI database as the interviews were being conducted. Finally, the data were weighted to account for the sample design and to reduce the effect of unit nonresponse. The population estimates used for post-stratification weighting were county-level estimates—grouped into region—by age and gender for July 1, 2003, published by the U.S. Census.
2005 Data Tables
The survey was conducted under contract to the Minnesota Department of Human Services by Clearwater Research, Inc. in Boise Idaho.
2001 Survey Methods
The sample consisted of households selected randomly from all telephone exchanges in each of four Minnesota regions ((1) Northeast --20 counties in the north central and northeastern part of the state; (2) West -- 33 counties along the western border of the state; (3) Southeast -- 24 counties northwest and southeast of the metro area; and (4) Twin Cities metro -- 9 counties). Selection of respondents occurred in two stages. First a household was randomly selected and then a person was randomly selected within the household. The initial screening was conducted to determine whether anyone in the household was eligible for participation in the survey.
The respondents: (1) were 55 years old or over (or 65 years old or over for the supplemental sample) and (2) had a permanent residence in Minnesota. When more than one person in the household met the criteria, the person who had the most recent birthday was selected to be interviewed.
The selection procedures guaranteed that every telephone household in each region had an equal chance to be included in the survey, and that once the household was sampled, every eligible adult had an equal chance to be included.
For each region, a total of 400 interviews were conducted with respondents age 55 and older. In addition, supplemental interviews were conducted with respondents age 65 and older to yield a regional total of 400 interviews with respondents 65 and older.
A total of 2,253 Older Minnesotans responded to the telephone interviews for the 2001 Survey of Older Minnesotans: 1,639 with respondents who were 55 or older, and 614 supplemental interviews with respondents who were 65 or older. The overall statewide response rate was 65 percent for the 55+ sample and 58 percent for the 65+ supplement sample.
“Weighting” the Data
All summary data presented have been “weighted” to accommodate the sampling frames: up-weighting the number of respondents under 65 years old, so that the true proportion of under 65 vs 65+ is maintained in each region; and correcting for regional oversampling, so that the statewide distribution reflects the actual distribution of the population 55 and older in Minnesota. Weighted response distributions will differ slightly from unweighted distributions. For example, although 2,253 actual interviews were completed, the statewide data file is based on 2,488 “weighted” cases.
The margin of error for a simple random sample of the size of the 2001 Survey of Older Minnesotans is plus or minus 2.0 percentage points, when the distribution of question responses is in the vicinity of 50 percent. The corresponding margin of error for each region’s weighted data file is plus or minus 4.2 percentage points for 550 completed interviews and plus or minus 4.0 percentage points for 600 completed interviews. These sampling errors presume the conventional 95 percent degree of desired confidence, which is equivalent to a “significance level” of .05. This means that no more than one time in 20 should chance variations in the sample cause the overall survey results to vary by more than 2.0 percentage points from the answers that would be obtained if all Minnesota residents 55 and older were interviewed.
2001 Data Tables
The 2001 Survey of Older Minnesotans was conducted by the Minnesota Center for Survey Research at the University of Minnesota. Major funding was for this survey was provided through the Minnesota Department of Human Services (PMQI Division). Additional funding was provided by Blue Cross/Blue Shield of Minnesota.