Comparing face-to-face to web data collection: unit response and costs in a national health survey (Preprint)

BACKGROUND Potential is seen in web data collection for population health surveys due to a combination of its cost-effectiveness, implementation ease and the increased internet penetration. Nonetheless, web modes may lead to lower and more selective unit response rates than traditional modes and hence may increase bias in the measured indicators. OBJECTIVE This research assesses the unit response and costs of a web versus F2F study. METHODS Alongside the F2F Belgian Health Interview Survey of 2018 (BHIS2018; n gross sample used: 7,698), a web survey (BHISWEB; n gross sample=6,183) is organized. Socio-demographic data on invited individuals is obtained from the national register and census linkages. Unit response rates considering the different sampling probabilities of both surveys are calculated. Logistic regression analyses examine the association between mode system (web vs. F2F) and socio-demographic characteristics on unit non-response. The costs per completed web questionnaire are compared with these for a completed F2F questionnaire. RESULTS The unit response rate is lower in BHISWEB (18.0%) versus BHIS2018 (43.1%). A lower web response is found among all socio-demographic groups, however, the difference is higher among people older than 65, low educated people, people with a non-Belgian nationality, people living alone and these living in Brussels Capital. Not the same socio-demographic characteristics are associated with non-response in both studies. Having another European (OR (95% CI): 1.60 (1.20-2.13)) or a non-European nationality (OR (95% CI): 2.57 (1.79-3.70)) (compared to having the Belgian nationality) and living in the Brussels Capital (95% CI): 1.72 (1.41-2.10)) or Walloon (OR (95% CI): 1.47 (1.15 - 1.87) region (compared to living in the Flemish region) is only in BHISWEB associated with a higher non-response. In BHIS2018 younger people (OR (95% CI): 1.31 (1.11-1.54)) are more likely to be non-respondent than older people, this was not found BHISWEB. In both studies, lower educated people have a higher change to be non-respondent, but this effect is more pronounced in BHISWEB (OR low vs. high education level (95% CI): Web 2.71 (2.21-3.39)); F2F 1.70 (1.48-1.95)). The BHISWEB study has a considerable cost advantage; the total cost per completed questionnaire is almost three times lower (€41) compared to the F2F data collection (€111). CONCLUSIONS The F2F unit response rate is generally higher, yet for certain groups the difference between web versus F2F is more limited. A considerable cost advantage of web collection is found. It is therefore worthwhile to experiment with adaptive mixed-mode designs to optimize financial resources without increasing selection bias; e.g. only inviting socio-demographic groups more eager to participate online for web surveys while remaining to focus on increasing the F2F response rates for other groups. CLINICALTRIAL Studies approved by the Ethics Committee of the University hospital of Ghent

Response patterns in the Belgian health interview survey: web versus face-to-face mode

Background Potential is seen in web data collection for population health surveys due to a combination of its cost-effectiveness and ease of implementation and the increased internet penetration rates. Nonetheless, web modes may lead to lower and more selective unit response rates than traditional modes such as face-to-face (F2F) interviewing and hence may increase bias in the measured indicators. This research assesses the response patterns of a web versus F2F study. Methods A Belgian health interview survey by web (BHISWEB; net sample=1010) was organized parallel to the traditional F2F BHIS 2018 (BHIS2018; net sample=11611; subsample used in this study=2748). Socio-demographic data on invited individuals were obtained from the national register and census linkages. To address the high item-missingness on education level deriving from the census, multiple imputation (m = 20) was applied. Unit response rates considering the different sampling probabilities of both surveys were calculated. Logistic regression analyses examined the impact of mode (web vs. F2F) and interactions between mode and socio-demographic characteristics on unit response. Results The unit response rate was significantly lower in the BHISWEB (18.0% (95% CI: 17.0-19.1)) than in the BHIS2018 (43.1% (95% CI: 41.5-44.7)). A lower web versus F2F response rate was found among all socio-demographic groups, however, this lower web response was less pronounced among youngsters, people cohabiting with household members, higher educated people and native Belgians. Conclusions The F2F unit response rate was generally higher, yet for certain groups the difference between web versus F2F was smaller. It is therefore worthwhile to experiment with adaptive mixed-mode designs to optimize resources without increasing selection bias; e.g. only inviting socio-demographic groups more eager to participate online for web surveys (e.g. youngsters) while remaining to focus on increasing the F2F response rates for other groups. Key messages The advantages of web versus F2F interviewing come against its considerable lower unit response rate. Differences in unit non-response between web and F2F vary between socio-demographic groups.

Muscle sympathetic single-unit response patterns during progressive muscle metaboreflex activation in young healthy adults

Muscle sympathetic single units respond differentially to sympathoexcitatory stress such that single units can increase firing to contribute to the sympathoexcitatory response or can be nonresponsive or even inhibited. We observed a subgroup of single units that can respond bidirectionally, being first inhibited before activated by progressive increases in forearm muscle metaboreflex activation. These results suggest convergent neural inputs (i.e., inhibitory and excitatory), which yield heterogenous muscle sympathetic single-unit activation thresholds.

Modeling of Low Impact Development Nutrient Reduction Performance in the Lake Simcoe Watershed

<p>Stormwater quality management has evolved from traditional centralized downstream control devices (e.g. ponds and wetlands) to distributed low impact development practices (LID) at the source (e.g. bioretention, porous pavement, greenroof).  In order to develop master LID plans for municipalities in the Lake Simcoe watershed (3576 km<sup>2</sup>), a new modeling approach was developed.  The challenge of modeling small scale LID practices over a watershed scale was resolved using unit response functions (URF) of different types of LID.  The concept of URF is based on the linear assumption of LID performance on a watershed level where routing is not important.  Detailed URF of runoff and nutrient reduction were developed on a lot level using US EPA SWMM models and linked with lot level characteristics such as imperviousness percentage.  The process of modeling include: (1) screening of appropriate LID across the watershed based on identification of unsuitable areas (e.g. wellhead protection area, NaCl concentration, industrial land use) and prioritization suitable lots which maximize environmental benefits and demonstration potential; (2) development of hydrological unit response functions of each type of LID (i.e. average annual runoff and nutrient loading reduction) using US EPA SWMM models; (3) aggregation of the cumulative runoff and nutrient reduction of all appropriate LID at each municipalities; (4) cost-effective analysis of different combinations of LID (i.e. Pareto front); (5) recommendation of the preferred LID combinations for each municipal within the watershed .  Results of the modeling indicate that (1) the average annual runoff volume reduction of implementing LID for the uncontrolled urban areas in Lake Simcoe watershed is estimated to be between 20% and 33%; and (2) the average annual phosphorus reduction of implementing LID for the uncontrolled urban areas in Lake Simcoe watershed is estimated to be between 2.0 to 2.7 tonnes per year.  This study has demonstrated a new modeling approach of small scale LID over watershed scales. </p>