Sensitive Questions in Online Surveys: An Experimental Evaluation of the Randomized Response Technique and the Crosswise Model

Self-administered online surveys provide a higher level of privacy protection to respondents than surveys administered by an interviewer. Yet, studies indicate that asking sensitive questions is problematic also in self-administered surveys. Because respondents might not be willing to reveal the truth and provide answers that are subject to social desirability bias, the validity of prevalence estimates of sensitive behaviors from online surveys can be challenged. A well-known method to overcome these problems is the Randomized Response Technique (RRT). However, convincing evidence that the RRT provides more valid estimates than direct questioning in online surveys is still lacking. A new variant of the RRT called the Crosswise Model has recently been proposed to overcome some of the deficiencies of existing RRT designs. We therefore conducted an experimental study in which different implementations of the RRT, including two implementations of the crosswise model, were tested and compared to direct questioning. Our study is a large-scale online survey (N = 6,037) on sensitive behaviors by students such as cheating in exams and plagiarism. Results indicate that the crosswise-model RRT---unlike the other variants of RRT we evaluated---yields higher prevalence estimates of sensitive behaviors than direct questioning. Whether higher estimates are a sufficient condition for more valid results, however, remains questionable.

Using Factorial Surveys and Stated Choice Experiments to Investigate Discriminatory Attitudes and Preferences

Empirical research on discriminatory attitudes and behaviour grapples with the social undesirability of its object. In many studies using regular survey methods, estimates are biased, and the social context of discrimination is not taken into account. Several methods have been developed, especially to deal with the first problem. In this regard, the estimation of the ‘true value’ of discriminatory attitudes is at the centre of interest. However, methodological contributions focusing on the social context of attitude communication and discriminatory behaviour, as well as the correlation between both, are rare. We present two experimental methods which address those issues: factorial surveys and stated choice experiments. In a first study, the usefulness of factorial surveys is demonstrated with data on German anti-Semitism (N=279). We show that the rate of approval with anti-Semitic statements increases if (a) respondents are told that the majority of fellows agree with such statements, (b) the term “Jews” is replaced by the term “Israelis”, and (c) reference to the Holocaust is made. Apart from the main effects of these experimental factors, significant interaction effects regarding the political attitudes and social status of respondents are observed. In a second study, a stated choice experiment on the purchase of olive oil and tomatoes was conducted in Germany (N=440). We find that respondents prefer Italian and Dutch products (control treatment) compared to Israeli and Palestinian ones (discrimination treatments). There are no significant differences between preferences for a so called ‘Peace product’ (which is produced jointly by Israelis and Palestinians) and products from Italy as well as the Netherlands. Yet, taking discriminatory attitudes (anti-Semitic and anti-Arabic attitudes) into account, a strong correlation between those attitudes and stated behaviour (purchase of Israeli, Palestinian and jointly produced products) can be found. This adds support to the hypothesis that discriminatory attitudes hold behavioural consequences.

Anreizformen und Ausschöpfungsquoten bei postalischen Befragungen. Eine Prüfung der Reziprozitätshypothese

Mittels eines Methodenexperiments mit zwei randomisierten Versuchsgruppen (N jeweils 200) und einer Kontrollgruppe (N=200) wird untersucht, ob und in welchem Ausmaß Geschenke bei einer postalischen Befragung zur Erhöhung der Ausschöpfungsquote beitragen. Es zeigt sich, dass ein versprochenes Geschenk (Telefonkarte im Wert von 10 Schweizer Franken) die Ausschöpfungsquote nicht erhöht, während ein dem Fragebogen beigelegtes Geschenk zu einem Anstieg der Quote um zirka 10 Prozentpunkte führt. Die Befunde stehen in Einklang mit der Reziprozitätshypothese, derzufolge Vorleistungen von vielen Personen auch dann honoriert werden, wenn die reziproke Handlung nicht dem unmittelbaren Eigeninteresse eines Akteurs entspricht.

Status and Future of Mountain Research: Results of two Surveys by the Research Initative of the Mountain Partnership

This report contains the results of two e-surveys conducted in 2005 and 2006 on the status and future of mountain research, as it presents itself within the Mt. Partnership. The surveys are the main outcome of a meeting of the leading members of the Mountain Partnership’s Research Initiative in 2004.

Erosive tooth wear and wedge-shaped defects in 1996 and 2006: cross- sectional surveys of Swiss army recruits.

The purpose of this study was to determine the prevalence and possible etiological factors of erosive tooth wear and wedge-shaped defects in Swiss Army recruits and compare the findings with those of an analogous study conducted in 1996. In 2006, 621 recruits between 18 and 25 years of age (1996: 417 recruits; ages 19 to 25) were examined for erosive tooth wear and wedge-shaped defects. Additional data was acquired using a questionnaire about personal details, education, dentitions subjective condition, oral hygiene, eating and drinking habits, medications used, and general medical problems. In 2006, 60.1% of those examined exhibited occlusal erosive tooth wear not involving the dentin (1996: 82.0%) and 23.0% involving the dentin (1996: 30.7%). Vestibular erosive tooth wear without dentin involvement was seen in 7.7% in 2006 vs. 14.4% in 1996. Vestibular erosive tooth wear with dentin involvement was rare in both years (0.5%). Oral erosive tooth wear lacking exposed dentin was also rare in those years, although more teeth were affected in 2006 (2.1%) than in 1996 (0.7%). The examinations in 2006 found one or more initial wedge-shaped lesions in 8.5% of the recruits, while 20.4% of the study participants exhibited such in 1996. In 1996, 53% consumed acidic foods and beverages more than 5 times/day; in 2006, 83.9% did so. In neither study did multivariate regression analyses show any significant correlations between occurrence and location of erosive tooth wear and wedge-shaped defects and various other parameters, e.g., eating and hygiene habits, or dentin hyper-sensitivity. Despite a significant increase in consumption of acidic products between 1996 and 2006, the latter study found both fewer erosive tooth wear and fewer wedge-shaped defects (i.e., fewer non-carious lesions.).

Training primary care physicians to offer their patients faecal occult blood testing and colonoscopy for colorectal cancer screening on an equal basis: a pilot intervention with before-after and parallel group surveys.

OBJECTIVES Primary care physicians (PCPs) should prescribe faecal immunochemical testing (FIT) or colonoscopy for colorectal cancer screening based on their patient's values and preferences. However, there are wide variations between PCPs in the screening method prescribed. The objective was to assess the impact of an educational intervention on PCPs' intent to offer FIT or colonoscopy on an equal basis. DESIGN Survey before and after training seminars, with a parallel comparison through a mailed survey to PCPs not attending the training seminars. SETTING All PCPs in the canton of Vaud, Switzerland. PARTICIPANTS Of 592 eligible PCPs, 133 (22%) attended a seminar and 106 (80%) filled both surveys. 109 (24%) PCPs who did not attend the seminars returned the mailed survey. INTERVENTION A 2 h-long interactive seminar targeting PCP knowledge, skills and attitudes regarding offering a choice of colorectal cancer (CRC) screening options. OUTCOME MEASURES The primary outcome was PCP intention of having their patients screened with FIT and colonoscopy in equal proportions (between 40% and 60% each). Secondary outcomes were the perceived role of PCPs in screening decisions (from paternalistic to informed decision-making) and correct answer to a clinical vignette. RESULTS Before the seminars, 8% of PCPs reported that they had equal proportions of their patients screened for CRC by FIT and colonoscopy; after the seminar, 33% foresaw having their patients screened in equal proportions (p<0.001). Among those not attending, there was no change (13% vs 14%, p=0.8). Of those attending, there was no change in their perceived role in screening decisions, while the proportion responding correctly to a clinical vignette increased (88-99%, p<0.001). CONCLUSIONS An interactive training seminar increased the proportion of physicians with the intention to prescribe FIT and colonoscopy in equal proportions.

Sensitive Questions in Online Surveys: Experimental Results for the Randomized Response Technique (RRT) and the Unmatched Count Technique (UCT)

Gaining valid answers to so-called sensitive questions is an age-old problem in survey research. Various techniques have been developed to guarantee anonymity and minimize the respondent's feelings of jeopardy. Two such techniques are the randomized response technique (RRT) and the unmatched count technique (UCT). In this study we evaluate the effectiveness of different implementations of the RRT (using a forced-response design) in a computer-assisted setting and also compare the use of the RRT to that of the UCT. The techniques are evaluated according to various quality criteria, such as the prevalence estimates they provide, the ease of their use, and respondent trust in the techniques. Our results indicate that the RRTs are problematic with respect to several domains, such as the limited trust they inspire and non-response, and that the RRT estimates are unreliable due to a strong false "no" bias, especially for the more sensitive questions. The UCT, however, performed well compared to the RRTs on all the evaluated measures. The UCT estimates also had more face validity than the RRT estimates. We conclude that the UCT is a promising alternative to RRT in self-administered surveys and that future research should be directed towards evaluating and improving the technique.

Sensitive Questions in Online Surveys: Experimental Results for the Randomized Response Technique (RRT) and the Unmatched Count Technique (UCT)

Gaining valid answers to so-called sensitive questions is an age-old problem in survey research. Various techniques have been developed to guarantee anonymity and minimize the respondent's feelings of jeopardy. Two such techniques are the randomized response technique (RRT) and the unmatched count technique (UCT). In this study we evaluate the effectiveness of different implementations of the RRT (using a forced-response design) in a computer-assisted setting and also compare the use of the RRT to that of the UCT. The techniques are evaluated according to various quality criteria, such as the prevalence estimates they provide, the ease of their use, and respondent trust in the techniques. Our results indicate that the RRTs are problematic with respect to several domains, such as the limited trust they inspire and non-response, and that the RRT estimates are unreliable due to a strong false "no" bias, especially for the more sensitive questions. The UCT, however, performed well compared to the RRTs on all the evaluated measures. The UCT estimates also had more face validity than the RRT estimates. We conclude that the UCT is a promising alternative to RRT in self-administered surveys and that future research should be directed towards evaluating and improving the technique.