A Comparison of Survey Techniques on Sensitive Sexual Behavior in Italy

Abstract

This article compares two national surveys carried out through the most commonly used procedures in Italy: CATI (computer-assisted telephone interviews) and SAQ-FI (self-answered questionnaires following interviews). Both surveys ask two identical questions concerning sensitive sexual behavior: early age at first intercourse and same-sex attraction. The SAQ-FI survey had both unit non-response and item non-response rates much lower than the CATI survey. Moreover, in the CATI survey, the groups with highest item non-response rates were also the groups with the lowest proportions of early intercourse and homosexual attraction. In addition, a differential analysis of the respondents produced diverse results for the two surveys. This is especially true of results by gender for same-sex attraction: Such behavior is more common among men (3.1%) than women (2.9%), according to the CATI survey, whereas the opposite is true of the SAQ-FI survey (6.1% of men vs. 7.7% women). In Italy at the beginning of the 21st century, CATI surveys reveal a lower level of early intercourse and same-sex attraction than SAQ-FI surveys. This article argues that the CATI survey underestimates the true level of these sensitive sexual behaviors in the Italian population.

Acknowledgments

This research was supported by the Italian Ministry for University and Scientific Research (Ministero dell'Università e della Ricerca Scientifica e Tecnologica).

Notes

¹As is commonly accepted in social science research, a unit non-response is the refusal on the part of a person extracted in a sample to participate in an interview, whereas a non-response or item non-response is the refusal on the part of a person who has agreed to participate in an interview to answer a particular question. In this article, we consider both aspects, which potentially generate distortion with respect to the objective of obtaining valid data for the entire sampling list (also known as the list of eligible people).

²Because the multipurpose computer-assisted telephone interviews survey (which already investigated several non-sensitive topics) could not be overly lengthy, we were only allowed to add two questions concerning sexual behavior. We chose the aforementioned questions for two reasons: (a) We expected to find the greatest discrepancies precisely between these two methods (especially in the Italian context), and (b) these two topics are among those most commonly investigated in sex-related research (Copas et al., 2002; Michaels & Giami, 1999).

³In 2006, only 24% of men aged 18 to 24 reported never having masturbated over the last year, compared to 66% of women. By contrast, 42% of men of the same age claimed to have masturbated at least once per week over the same period, compared to just 14% of women (Barbagli, Dalla-Zuanna, & Garelli, 2010).

⁴We divided Italy into two macro-areas: Northern regions (those located north of the capital city, including Rome) and Southern regions (those south of Rome). These two areas have very different demographic, social, and economic characteristics: Northern regions are more similar to other Western European countries, whereas Southern regions are more similar to Mediterranean European countries.

⁵For the 2004 Survey of Health, Ageing, and Retirement in Europe, see http://www.share-project.org/. For the 2008 Bank of Italy survey on Household Income and Wealth, see http://www.bancaditalia.it/statistiche/indcamp/bilfait/boll_stat/en_suppl_08_10.pdf (p. 40).

⁶On the Autorità per le Garanzie nelle Comunicazioni ‘Italian Authority for Communications’ institutional Web site, the Italian survey company must declare a number of parameters concerning the surveys they perform, including the unit response rate (see http://www.agcom.it/).

Note. The reference group is in parentheses. SAQ-FI = self-answered questionnaires following interviews; CATI = computer-assisted telephone interviews. LR = likelihood ratio.

*p < .10. **p < .05. ***p < .01.

⁷The binary logistic (logit) regression is a nonlinear regression used when the dependent variable is binary (i.e., a categorical variable that has two values, such as yes/no or present/not present). It estimates the natural logarithm of the probability (odds) that an event happens on the basis of continuous and categorical covariates (Agresti, 1996). When reading the results of a logit model, an estimated odds ratio significant and >1.0 means that respondents with a certain characteristic (e.g., being male) are more likely to have the value of interest of the dependent variable (e.g., having experienced early intercourse) than the respondents who have the characteristics used as references (e.g., being female).

Note. The reference group is in parentheses. SAQ-FI = self-answered questionnaires following interviews; CATI = computer-assisted telephone interviews LR = likelihood ratio.

*p < .10. **p < .05. ***p < .01.

Note. N = 3,058.