Introduction
These data notes provide you with information on the Rapid Risk Factor Surveillance System (RRFSS), as well as methods and limitations associated with RRFSS health indicator reports produced by the Halton Region Health Department. View other health indicator reports.
About RRFSS
The Rapid Risk Factor Surveillance System is an on-going telephone survey (land line and cell phone) used to collect information on attitudes, behaviours, knowledge and awareness of issues related to health in Halton. RRFSS is conducted by the Institute of Social Research at York University.
Since 2001, each year a random sample of approximately 1,200 adults aged 18 and over are surveyed in Halton Region. The information collected in these surveys is used by the Halton Region Health Department to produce health indicator reports which support program planning and evaluation, policy development, and help to improve awareness of health issues in the community.
Weighting
Analysis of RRFSS data uses dual-frame sampling weights to account for the two different sampling frames (landline and cellphone) and for the probability of being selected to complete the survey. For instance, the weights adjust for the fact that in landline sample, an adult from a household with a large number of adults is less likely to be selected to complete the survey than an adult who lives alone.
Making comparisons
In the health indicator reports comparisons are typically done by sex, age, municipality, income and education for questions related to individuals, and by municipality and income for questions related to households. Comparisons over time are also presented when multiple years of data are available.
It is not recommended to compare RRFSS data from 2016 onwards to previous RRFSS data due to changes in sampling and analysis methodology. For example, beginning in 2016 a cell phone sample was included and surveyors started asking to speak to individuals aged 18-30 first in order to boost the number of young adults included in the survey sample.
Calculating income groups
Income groups are determined by first asking respondents about their household income. If respondents refuse to provide their household income they are then asked to provide a range of income and the midpoint of the range is used to estimate household income. Second, the respondent’s adjusted household income is calculated by dividing household income by the square root of their household size. Adjusted household income reflects the fact that a household’s needs increase as the number of members increase. The adjusted household income for all Halton respondents are then organized into 10 equal deciles and placed into low (decile 1-3), middle (decile 4-7) and high (decile 8-10) income groups. Respondents who do not know or refuse to provide their household income (about 15% of respondents) are excluded from the income analysis.
Statistical significance and variation
Overlapping 95% confidence intervals (CIs) are used to determine statistical significance in health indicator reports. A 95% confidence interval refers to a range of values that have a 95% chance of including the true estimate. When CIs do not overlap between 2 or more groups (e.g. when comparing males and females) it means that the differences between the groups are statistically significant and unlikely to be due to chance alone. Since overlapping confidence intervals are used to determine statistical significance, p-values are not calculated. This is a conservative approach which is more appropriate when multiple comparisons are being made, such as in health indicator reports. CIs are used to determine statistical significance, however CIs are not always presented in health indicator reports.
Coefficient of variation (CV) refers to the precision of an estimate. When the CV is between 16.6 and 33.3, the estimate should be interpreted with caution because of high variability, and has been marked with an asterisk (*). Estimates with a CV of greater than 33.3 are not reportable and have been marked with double asterisks (**) in the figures and tables.
Limitations
RRFSS results are self-reported and may not be recalled accurately. Individuals not living in households (such as those in prison, hospitals, or the homeless) are excluded. As a result, the percentages may not represent the true estimates for the general population. In Halton, the survey is administered in English only.
Rounded estimates are used for the presentation of data, therefore estimates may not total 100%.
Don’t know and refused responses are typically excluded from the analysis. When “don’t know” is considered a valid response, or when over 5% of respondents answer “don't know”, the response is included in the analysis.
Some analyses are limited by sample size.
Additional resources