Search

Objective: This study examines the effectiveness and costs of follow-up phone calls in improving response rates to a community survey. Methods: Non-responders to a postal survey were randomly allocated to receive a phone call or no phone call. The resources used for the development and implementation of the survey were documented. The response rates and cost per level of follow-up contact examined. Results: Follow-up phone calls led to a statistical significant increase in the number of responses to a community-wide survey, relative to no phone call. This relative increase in responses (n=62 for the follow-up phone call group versus n=l for controls), did not increase the absolute survey response rate sufficiently (from 38.5% for two mailed surveys to 39.8% for two mailed surveys plus a phone call) to justify the phone call costs. Scenario analyses show increasing the initial response rate by 10% and conducting a second mailed survey achieves greater marginal cost savings than increasing the response rate to the second mailout or the follow-up phone calls. Conclusions: These results suggest a follow-up phone call was not cost effective. Survey research ought to primarily focus on obtaining optimal initial response rates by using strategies identified in a Cochrane meta-analytic review. Key words: response rate, non-response, survey, follow-up, cost, randomised controlled trial. Aust NZ J Public Health. 2010; 34:508-12 doi: 10.1111/j.1753-6405.2010.00598.x

The questions of whether it is worth using monetary incentives and the size of the inducement by testing a regression model of the impact of prepaid monetary incentives on response rates in consumer and organizational mail surveys is tackled. The results support their use and the inducement value that makes a significant impact on the effect size is shown.

The timing of follow-up efforts with regard to mail survey research is examined.

Increasing use of incentive payments to survey respondents raises the threat of several unintended consequences, among them the creation of expectations for future payments and the possibility of a deterioration in the quality of response. Such deterioration may come about either as a direct result of substituting external for internal motivation, or as a consequence of expectations for rewards that go unmet by the survey organization. The findings from the present study are somewhat reassuring with respect to both of these unintended outcomes. Although people who have received a monetary incentive in the past are significantly more likely than those who have not to endorse the statement that 'people should be paid for doing surveys like this,' they are actually more likely to participate in a subsequent survey, in spite of receiving no further payments. And respondents who received an incentive 6 months earlier are no more likely than those who received no incentive to refuse to answer (or to answer 'don't know' to) a series of 18 key questions on the survey. Furthermore, they are more likely than other respondents to express favorable attitudes toward the usefulness of 'surveys like this.' The generality of these findings, however, needs much further testing., There is some evidence that the difficulty of obtaining cooperation with sample households in the United States and other developed countries is growing over time (de Heer and Israels 1992). [...]

Objective: Much attention is paid to the influence of different data collection methods on the quality of self-reported drinking behavior estimates. Thus far, however, the findings show inconsistencies. Therefore, a comprehensive study was conducted to compare data on alcohol use and alcohol-related problems obtained by mail survey and personal interviews. Method. A general population survey on alcohol was conducted among a random sample of 8,000 Dutch inhabitants of Rotterdam aged 16 to 69. A small sample (n = 500) of the total sample (N = 8,000) was personally interviewed and the others (n = 7,500) received a mailed questionnaire. The response rate was 44% (N = 3,537). Respondents of the mail survey and personal interviews are compared on overall response rate, item-nonresponse rate, background factors, self-reported alcohol use, alcohol-related problems and problem drinking. Results: No notable differences in self-reported alcohol use, alcohol-related problems or problem drinking were found by data collection mode. This holds for both the total general population and for men and women separately. The overall response rate was somewhat higher for the personal interviews. No important significant differences were found in item nonresponse or background factors. Conclusions: The absence of notable differences in estimated self-reported drinking behavior by mail survey and personal interviews indicates that both data collection methods yield comparable results. This is true for both the total population and for men and women separately. (J. Stud. Alcohol 59: 280-285, 1998), Researchers conducted a study to determine and compare data on alcohol use and alcohol-related problems obtained by personal interviews and mail. They found out that there is little difference in results obtained whether by mail surveys or personal interviews. This holds true both for the total general population and for women and men separately., Scientific knowledge is based on weighing and considering the findings of several studies. This analysis, however, is complicated by the notion that inconsistencies and even consistencies may be due to [...]

Introduction Kazuaki (1990) pointed out that market research clients nowadays are more sensitive to cost and speed than they used to be. To satisfy these requirements, the global emergence of [...]

This study reports results of an experiment conducted to compare response rates and response content in mail and electronic mail surveys. Respondents on a large global network (Internet) were sent [...]

This article reports the results of an experiment which examined the effect of redeemable discount coupons on rate of response and cost per response in a mail survey conducted among [...]

The aim of this note is to describe the procedure followed by the authors in solving a problem in mail questionnaires in which there were questions where the respondents were [...]

For several decades researchers have been concerned with non-response and response biases in mail surveys (Suchman & McCandless 1940; Toops 1926). The present study examines the effect that sample members' [...]

Organizational surveys often have fairly low response rates. Sample surveys with low response rates can produce biased samples, particularly if key organizational characteristics affect the pattern of survey response. In [...]

Research using a survey of registered voters' assessment of a local government provides evidence of the importance of using special handling mail survey procedures. Late respondents held different opinions on government activities than early respondents and were also more likely to be less educated. Differences between early and late respondents remained even when controlling for education. The findings support the belief that important benefits are derived from the additional costs incurred for special mailings.

Although data collection in conjoint analysis has traditionally involved the use of in-person interviews, recent years have seen a trend toward the use of alternative data collection modes, including mail questionnaires and telephone interviews. Since these alternative modes differ in environment, the author examines the predictive performance of conjoint models under three data collection modes, i.e., in person interviews, mail questionnaire, and telephone interviews. The results indicate the conjoint models examined to be comparable in predictive performance across the three data collection modes. (Reprinted by permission of the publisher.)

This study assessed differences in response rate when the author of a survey was identified as a graduate student or a faculty member. A survey on research ethics was mailed to 500 nurse faculty members with half of the covering letters signed as a graduate student and half signed as a faculty member. There were 388 (78%) out of a possible 495 surveys returned. The response rate was 78% for graduate student and 77% for faculty identifications, not significant by chi-square test.

The research shows that consumer anonymity has only small effects on response variables for mail surveys. Respondents who were anonymous gave a little less information to complicated questions; respondents in the lower-income class, homeowners, and non-whites were more willing to provide their identity on the return envelope. Nevertheless, the information provided by these groups was not that different from the information provided by the other respondents for a range of questionnaire items evaluated.