The 1996/97 Ontario Health Survey
The 1990 Ontario Health Survey
Comparison of the 1990 and 1996/97 Ontario Health Surveys The 1996/97 Ontario Health Survey (OHS)
Original source: Statistics Canada
Distributed by:
1. Knowledge Management and Reporting Branch, Ontario MOHLTC
2. Statistics Canada
Suggested citation (see Data Citation Notes):
1. Ontario Health Survey 1996/97, Statistics Canada, Share File, Knowledge Management and Reporting Branch, Ontario MOHLTC
2. Ontario Health Survey 1996/97, Statistics Canada, Public Use Microdata File, Statistics Canada Data Notes - The 1996/97 OHS was conducted from October 1996 to August 1997 in conjunction with the National Population Health Survey (NPHS). The Ontario response rate to the NPHS was 78.8% of households (general survey) with 94.4% of selected individuals subsequently completing the health component. The 1996/97 OHS is comprised of 94.5% of these individuals aged 12+ who agreed to share their information with the province. The Ontario buy-in was conducted using computer-assisted telephone interviewing with sampling by random digit dialing. The buy-in supplemented the core NPHS sample. All health units were sampled and grouped into 23 health areas.
- The 1996/97 OHS excluded certain groups such as the homeless, aboriginal people on reserves, and those living in institutions.
- Response rates varied by geographical area.
- Because of the complex sampling techniques, the 1996/97 OHS requires that sampling weights be used when tabulating data.
- Estimates with a coefficient of variation greater than 33.3% or based on cell sizes of less than 30 responses must be suppressed. This can be a common occurrence for some areas, particularly when the data are sub-divided by age and sex. Estimates with a CV between 16.6% and 33.3% should be noted as having high sampling variability.
- Depending upon the question, data may be subject to recall bias, social desirability bias and errors from proxy reporting. Few OHS questions were answered by proxy.
- The OHS is often the only source for certain types of health data; it covers the general population, and focuses on broad aspects of health, including lifestyle behaviours, health problems, and health care utilization.
References and Resources - Ontario Ministry of Health and Long-Term Care. 1996-97 NPHS: Ontario Health Survey File. Record Layout. April 1999.
- Statistics Canada. National Population Health Survey Overview 1996/97. Ottawa: Ministry of Industry, 1998. Catalogue no. 82-567-XPB.
The 1990 Ontario Health Survey (OHS)
Original source: Ontario Ministry of Health (renamed Ministry of Health and Long-Term Care)
Distributed by: Knowledge Management and Reporting Branch, Ontario MOHLTC
Suggested citation (see Data Citation Notes):
Ontario Health Survey 1990, Ontario MOHLTC Data Notes - The 1990 OHS was conducted from January to November, 1990 (excluding July); the Ontario response rate was 87.5% for the interviewer-completed questionnaire and, of those with an interview, 77.2% returned the self-completed questionnaire. The interviews were face-to-face. All 42 public health units that existed at the time were sampled.
- The 1990 OHS excluded certain groups such as the homeless, aboriginal people on reserves, and those living in institutions.
- Response rates varied by geographical area.
- Because of the complex sampling techniques, the 1990 OHS requires that sampling weights be used when tabulating data.
- Percentages with a coefficient of variation (CV) greater than 25.0% or based on cell sizes of less than 30 responses must be suppressed. This can be a common occurrence for some areas, particularly when the data are further sub-divided by age and sex. Percentages with a CV greater than 16.6% should be qualified.
- Depending upon the question, data may be subject to recall bias, social desirability bias and errors from proxy reporting. Few OHS questions were answered by proxy.
- The OHS is often the only source for certain types of health data; it covers the general population, and focuses on broad aspects of health, including lifestyle behaviours, health problems, and health care utilization.
References and Resources - Ontario Ministry of Health. Ontario Health Survey 1990: Highlights Report. Toronto: Ontario Ministry of Health, 1991.
- Ontario Ministry of Health. Ontario Health Survey 1990: User’s Guide Volume 1, Documentation. Toronto: Ontario Ministry of Health, 1991.
Comparison of the 1990 and 1996/97 Ontario Health Surveys | | 1990 OHS | 1996/97 OHS | | Geographical unit | 42 areas: public health units | 23 health areas: health unit groupings | | Ontario sample size | 49,164 (Ages 12+) | 36,892 (Ages 12+) | | Target sample size per area | 1,000 per PHU | 1,200 per area | | When survey was conducted | January-November 1990, excluding July | October 1996 to August 1997 | | Response rates | Household – 87.5%
O5 – 77.2%
Overall – 67.6% | Households – 78.8%
Selected – 94.4%
Sharing – 94.5%
Overall – 70.3% | | Methodology and implications | Personal face-to-face interview with interviewers completing the household questionnaire (03/04 form) and respondents filling in a self-completed questionnaire (05 form). The O3/O4 was answered by a knowledgeable member of the household on behalf of everyone in the household and the O5 was self-completed by everyone in household. There was higher non-response on the self-completed questionnaire than interview questionnaire. The self-completed questionnaire would have been problematic for those with limited literacy skills, although interviewers did provide assistance if needed. | Some personal interview but mostly conducted by telephone using computer-assisted telephone interviewing (CATI). The General File was answered by a knowledgeable member of the household (data only on NPHS, not OHS) and the Health file was completed by interview (mostly telephone) of a randomly-selected individual in household. Random selection resulted in less clustering, lower design effect, and smaller coefficient of variation (c.v.). The CATI format resulted in fewer item non-response, allowed for "up front" edits that reduced error. | | Release Guidelines | Estimates with a C.V. of 0 -16.5% can be released with no restriction, C.V. of 16.6 – 25.0% used with qualification, C.V. greater than 25.0% should not used at all. | Estimates with a C.V. of 0-16.5% can be released with no restriction, 16.6 - 33.3% used with qualification, C.V. greater than 33.3% should not used at all. |
|
