Why were there three? – Determinants of the presence of an intimate partner during face-to-face interviews

---

Abstract

This study analyses determinants of the presence of an intimate partner during face-to-face interviews. Based on theoretical assumptions about opportunity structure, social control, social support, and companionship, we investigated partner presence using data from the first wave of the German Family Panel (pairfam). Descriptive results revealed that an intimate partner was present in every seventh interview. Multivariate results using separate logistic regression models for the presence of the female (n = 3,272) and the male partner (n = 2,348) revealed that the opportunity structure, such as the couple’s living arrangements or their employment status, had the greatest influence on the presence of both female and male partners. Gender differences existed for social control, social support and companionship. The results suggest that partner presence could most easily be prevented by taking into account the opportunity structure.

Keywords

interview characteristics, intimate partner, survey management, third person

Introduction

Face-to-face interviews are influenced by aspects of both interviewer (Cannell et al. 1977) and respondent (Billiet & Davidov, 2008) as well as by characteristics of the interview situation (Johnson 2014) that may result in systematically over- or underestimated measures of survey variables (Krumpal 2013). A special influence on the interview situation is the presence of a third person during the interview. Earlier research found that the third person most likely to be present during an interview is the intimate partner (Aquilino 1993; Hartmann 1994; Lander 2000). Previous research examining the presence of the intimate partner detected differential measurement outcomes of partnership related questions, such as marital dissolution, self-reported levels of marital conflict (Aquilino 1993) and partnership satisfaction (Mohr 1986).

Determinants influencing the presence of the intimate partner that are discussed in the sparse literature include structural (Reuband 1992) as well as motivational reasons or personality characteristics of the respondent or the partner (Reuband 1987; Hartmann 1994; Lander 2000). Empirical research found that non-working status was associated with a higher chance of partner presence during the interview and socioeconomic status was inversely associated with partner presence (Aqulino 1993). Sex differences between the interviewer and the respondent were associated with a higher chance of partner presence when a woman was interviewed by a man, and this was interpreted as social control (Hartman 1994). Although the intimate partner is most likely to be the third person present during the interview, an examination of relationship quality and presence of the intimate partner is lacking in the literature. However, the presence of the intimate partner during the interview is highly likely to bias the results when both partners are concerned about potential survey questions; hence the respondent tends towards socially motivated misreporting (Aqulino et al. 2000; Krumpal 2013). The intimate partner could, however, also assist the respondent emotionally or provide important information during an interview. Partner presence may thus also have positive effects (Reuband 1984). Both are to be expected in the German Family Panel (pairfam) (Huinink et al. 2011), because the major research issues of pairfam are couple dynamics and partnership stability. Pairfam addresses the development of partnership relations, including the quality of the relationship, issues of labour division, the internal power distribution and the stability of the relationship in a multi-actor design, in which information on both the respondent and the partner are available.

The aim of the present study is to extend the existing literature on the presence of the intimate partner and investigate the association between characteristics of the intimate relationship and the presence of the intimate partner. This paper does not investigate the potential influences of partner presence on the response behaviour of the respondent. The paper starts with brief theoretical considerations, introduces data and method, presents the results and discusses the findings in the conclusion.

 

Theoretical background and hypotheses

The opportunity structure should influence the partner’s presence. The employment status or living arrangements are crucial for partner presence (Hartmann 1994). The chance that the partner will be present should be more likely if both partners live in the same household. In contrast, living in separate households should be a higher barrier for partner presence because it is more difficult to schedule a joint appointment. The partner might also be present because he is not employed and has time to spend with the respondent.

H1: The partner is more likely to be present if both partners live in the same household.

H2: The partner is more likely to be present if the partner is not employed.

In line with social control theory, people try to gain control over situations to influence these in a beneficial way, i.e. seek desired events and avoid undesirable ones (Thompson 1981). Accordingly, the partner should be more likely to be present if the partner is keen to know the respondent’s answers in an interview, so as to influence the response behaviour in a beneficial way. A more positive interpretation of this potential motive is curiosity. But it is reasonable to assume that control motivation is a characteristic of mistrust and thus of a disturbed relationship. Control motivation should therefore be more likely if the partner is dominant. Another reason for a higher social control motivation might be the sex constellation between the respondent and the interviewer (Hartmann 1994). Especially in situations in which the respondent and the interviewer are of opposite sex the partner might be jealous and curious what will happen during the interview. Furthermore, jealousy could be a reason for a higher control motivation if the respondent is very attractive (Lander 2000; Guerrero et al. 2004).

H3: The higher the partner’s control motivation, the higher the chance of the partner’s presence during the interview.

In line with the social support theory, a further assumption is that the interview situation might be regarded as a stressful event for the respondent, in particular, if a respondent is being interviewed for the first time (Lakey & Cohen 2000). The respondent might feel uncomfortable with this unknown situation. Therefore the partner might be more likely to be present to support the respondent during the interview. This should occur in partnerships that are characterised by strong reciprocal support. Furthermore, the presence of the partner as a source of support might be desired if the respondent has low self-worth and is uncertain how to answer survey questions. Lander (2000) mentioned that a higher educational level of the partner could be an indicator of social support. Thus, a higher educational level of the partner should increase the chance of his presence.

H4: The higher the partner´s social support, the higher the chance that the partner will be present during the interview.

H5: The lower the respondent´s self-worth, the higher the chance that the partner will be present.

H6: A higher educational level of the partner increases the chance of partner presence.

In contrast to the partner’s control motivation, companionship might influence the partner’s presence during the interview (Aquilino 1993). A higher level of trust and intimacy within the relationship is likely to enhance the partner’s presence. Partner presence might be more likely if the partners share their secrets and their leisure time activities. Companionship is expected to be a characteristic of good relationships.

H7: The greater the intimacy between the two partners, the higher the chance that the partner will be present during the interview.

 

Data and method

Pairfam enables extended empirical research on issues of couples and family development with a focus on couple dynamics and partnership stability, childbearing, parenting and child development, and intergenerational relationships (Huinink et al. 2011). Although pairfam’s main focus is couples and family development, the paradata (information about the process of data collection) allow methodological questions to be investigated, such as the presence of the intimate partner during the interview. The annual survey is based on a cohort-sequence design and started in 2008 with 12,402 randomly selected respondents from the three birth cohorts 1991-93, 1981-83, and 1971-73. Population registers were used as the sampling frame.

To assess determinants of partner presence during the interview, we use data from the first wave of the pairfam study collected in 2008/09. The overall response rate for the first wave was 36.9%. This relatively low response rate is common for German surveys and does not result in a nonresponse bias (Huinink et al. 2011). Our analysis includes all three birth cohorts and is restricted to persons with an intimate heterosexual relationship. In pairfam, respondents are interviewed with a Computer-Assisted Personal Interview (CAPI). Sensitive questions or questions about relationship quality which might be regarded as sensitive in the presence of other household members are asked by Computer-Assisted Self-Administered Interview (CASI). The interviewer´s laptop is handed over to the respondent and the respondent fills out the questions autonomously (Huinink et al. 2011: 92).

Operationalisation

The binary dependent variable partner presence during the interview (partner is present: 1; partner is not present: 0) is based on paradata recorded by the interviewer after each successful interview. The interviewer had to specify whether the partner, children or other persons were present during the interview or not.

Respondent´s attractiveness (very unattractive: 1; very attractive: 7) is based on the interviewer´s assessment. All other information including information about the partner´s employment status and educational level were provided by the respondent.

Opportunity structure is indicated by the respondent´s and the respondent´s partner employment status (respondent/partner was not employed (including unemployment, retirement and parental leave): 1; full- or part-time employment: 0) as well as the cohabitation status (partners live in the same household: 1; partners live in separate households: 0).

Dominance of the partner, the interviewer-respondent sex constellation and the above-mentioned attractiveness of the respondent were considered as indicators of control motivation. Dominance within the respondent-partner dyad (low: 1; high: 5) was measured via an adapted and shortened version of the Network of Relationships Inventory (NRI) (Furman and Buhrmester 1985). The respondent reported how often his/her partner gets his/her way when they can´t agree on something and how often their partner makes them do things in his/her way. These questions were answered in CASI mode. The interviewer-respondent sex constellation is coded 1 if the respondent and interviewer have the opposite sex and 0 for the same sex.

Social support is operationalised via dyadic coping within the respondent-partner dyad and the self-worth of the respondent. The measurement of dyadic coping (low: 1; high: 5) is based on the supportive dyadic coping of the partner scales from the Dyadic Coping Questionnaire (FDCT-N) (Bodenmann 2000). The respondent answered several items about how often and in which way the partner supports him or her in stressful situations. The respondent’s self-worth (low: 1; high: 5) was measured via three items from an adaptation of Rosenberg’s Self-Worth scale (1965). Items for dyadic coping and self-worth were retrieved via CASI mode. The respondent´s and their partner´s educational level were measured via the International Standard Classification of Education (ISCED) (Schneider 2008). A variable comparing the educational level of the two partners was included in the analysis. This variable indicates whether the respondent and the partner have the same educational level or whether the partner or the respondent is more highly educated.

Companionship was operationalised by the measure of intimacy within the respondent-partner dyad (low: 1; high: 5) and was also collected via CASI mode. The respondent answered items from the Intimacy Network of Relationships Inventory (NRI) (Furman & Buhrmester 1985) on how often he shares his feelings and secrets with his partner.

Relationship duration in months, respondent´s age, partner´s age, and birth cohort were included as control variables in the analysis. Table A2 in the appendix lists detailed information about the scales and items used.

Statistical analysis

Descriptive statistics were calculated to reveal the prevalence of partner presence in the first wave of pairfam. As social support and control motivation may have different effects, depending on the sex of the partner (Lander 2000), separate logistic regression models for the presence of the male partner (n = 3,272) and presence of the female partner (n = 2,348) with average marginal effects (AMEs) were computed to allow the stepwise inclusion of variables and the comparison of coefficients across models (Long & Freese 2014, Mood 2010). As interviews were nested within interviewers (minimum interviews per interviewer was one, maximum was 85) and rates of partner presence differ by interviewer, we accounted for a potential clustering with robust standard errors using interviewers as clusters.

 

Results

Descriptive Results

Table 1 shows descriptive results regarding partner presence as a function of birth cohort, living arrangement and respondent’s sex. In 14.88% of interviews, an intimate partner was present. Stratified by the three birth cohorts, partner presence was 3.83% for the 1991-1993 birth cohort, 15.74% for the 1981-1983 birth cohort and 15.67% for the 1971-1973 birth cohort (Chi² = 40.63, p = 0.00). A higher degree of institutionalisation of the relationship was associated with a higher percentage of partner presence (living in separate households = 5.02% vs. cohabitation = 15.81% vs. marriage = 18.65%; Chi² = 132.65, p = 0.00). Partner presence was higher for male respondents than for female respondents (19.25% vs. 11.74%, Chi² = 60.95, p = 0.00).

Table 1 Percentage of partner presence stratified by birth cohort, living arrangement and respondent’s sex

Respondents with partner	14.88
Birth cohort
1991-1993	3.83
1981-1983	15.74
1971-1973	15.67
Living arrangement
Living in separate households	5.02
Cohabitation	15.81
Marriage	18.65
Respondent’s sex
Male (female partner present)	19.25
Female (male partner present)	11.74
N	5620
Source: pairfam wave 1 2008/2009;calculations by the authors

 

Multivariate Results

Table 2 shows stepwise logistic regression models for the probability of the presence of the female partner during the interview of a male respondent. Model 1a includes only the opportunity structure and the control variables. If both partners lived in the same household the probability of the presence of the female partner was 17.2% higher than for couples living in separate households. Furthermore, the chance that the female partner would be present during the interview increased by 15.5% if the male respondent was unemployed. If the female partner was unemployed the chance increased as well. In Model 2a variables for social control were added. The results show a significant positive effect for an opposite interviewer-respondent sex constellation: female partners were more often present (AME = 5.5%) during the interview if a male respondent was interviewed by a woman. Surprisingly, the more attractive the male respondent was, the lower the chance was that the female partner was present. Differences in the educational level were not associated with partner presence.

In Model 3a variables for social support were added. The lower the self-worth of the male respondent, the higher was the probability of the presence of the female partner. Intimacy added in Model 4a had a significant positive effect of 3.1% on the presence of a female partner. A female partner was more often present during the interview if the intimacy between the male respondent and the female partner was high. Nagelkerkes R² slightly increases from Model 1a to Model 4a (0.11 vs. 0.13).

Table 2 Logistic regression with AMEs (with logits in brackets) for the presence of a female partner during an interview of a male respondent

	Presence of female partner
	Model 1a		Model 2a		Model 3a		Model 4a
	AME	Logit	AME	Logit	AME	Logit	AME	Logit
Opportunity structure
Living in the same household
No	ref		ref		ref		ref
Yes	.172***	(1.652)	.173***	(1.671)	.174***	(1.685)	.174***	(1.686)
Male respondent not employed
Employed	ref		ref		ref		ref
Not employed	.155***	(.887)	.147***	(.855)	.143***	(.836)	0.141***	(.831)
Female partner not employed
Employed	ref		ref		ref		ref
Not employed	.083***	(.529)	.082***	(.530)	.082***	(.531)	.081***	(.527)
Social Control
Dominance of female partner			-.009	(-.062)	-.011	(-.076)	-.014	(-.097)
Interviewer-respondent sex constellation
Male respondent – male interviewer			ref		ref		ref	ref
Male respondent – female interviewer			.055*	(.375)	.055*	(.379)	.055*	(.382)
Attractiveness of male respondent			-.018*	(-.127)	-.018*	(-.122)	-.017*	(-.117)
Social support
Dyadic coping of female partner					.020	(.137)	.004	(.025)
Self-worth of male respondent					-.022+	(-.154)	-.025*	(-.177)
Educational level of respondent and partner
Same educational level					ref		ref
Male respondent has higher education					.003	(.024)	.003	(.020)
Female partner has higher education					-.003	(-.021)	-.004	(-.026)
Companionship
Intimacy							.031**	(.219)
Control variables
Relationship duration	-.001	(-.001)	-.001	(-.001)	-.001	(-.001)	-.001	(-.000)
Age of respondent and partner
Same age	ref		ref		ref		ref
Male respondent is older	-.024	(-.172)	-.026	(-.190)	-.026	(-.190)	-.025	(-.185)
Female partner is older	.009	(.058)	.005	(.037)	.005	(.035)	.005	(.034)
Birth cohort
1991-1993	ref		ref		ref		ref
1981-1983	.132*	(1.328)	.132*	(1.124)	.131*	(1.334)	.135*	(1.141)
1971-1973	.128*	(1.296)	.130*	(1.335)	.129*	(1.317)	.133*	(1.394)
N	2,348		2,348		2,348		2,348
Nagelkerke R2	.110		.123		.126		.131
Source: pairfam wave 1 2008/2009; calculations by the authors; average marginal effects (AME); logit coefficients in parenthesis; robust standard errors for interviewer ID; ***p<0.001; **p<0.01; *p<0.05; +p<0.1

Table 3 shows the results for the presence of the male partner. In comparison to the presence of the female partner, the AMEs for the opportunity structure variables have the same sign but are smaller in size. In contrast to the presence of the female partner, the analysis did not reveal any significant AMEs for social control. In particular, if a female respondent was interviewed by a male interviewer the chance that the male partner would be present did not increase (Model 2b). The lower the self-worth of the female respondent, the higher was the chance of the male partner´s presence. Contrary to the presence of the female partner, dyadic coping (Model 3b) was positively associated with the presence of the male partner. Compared with the presence of the female partner Model 4b does not reveal an effect of intimacy on the presence of the female partner. Nagelkerkes R² increased marginally from Model 1b to Model 4b (0.057 vs. 0.064).

Table 3 Logistic regression with AMEs (with logits in brackets) for the presence of a male partner during an interview of a female respondent

	Presence of male partner
	Model 1b		Model 2b		Model 3b		Model 4b
	AME	Logit	AME	Logit	AME	Logit	AME	Logit
Opportunity structure
Living in the same household
No	ref		ref		ref		ref
Yes	.079***	(.987)	.079***	(.982)	.081***	(1.018)	.081***	(1.020)
Female respondent not employed
Employed	ref		ref		ref		ref
Not employed	.058***	(.540)	.057***	(.527)	.055***	(.516)	0.055***	(.514)
Male partner not employed
Employed	ref		ref		ref		ref
Not employed	.060**	(.512)	.059*	(.502)	.056*	(.485)	0.056*	(.482)
Social Control
Dominance of male partner			.004	(.042)	.004	(.042)	.0040	(.042)
Interviewer-respondent sex constellation
Female respondent – female interviewer			ref		ref		ref	ref
Female respondent – male interviewer			.011	(.108)	.011	(.113)	.011	(.115)
Attractiveness female respondent			-.003	(-.030)	-.002	(-.020)	-.002	(-.020)
Social support
Dyadic coping of male partner					.020*	(.195)	.021*	(.212)
Self-worth of female respondent					-.018*	(-.181)	-.018*	(-.177)
Educational level of respondent and partner
Same educational level					ref		ref
Female respondent has higher education					.001	(.013)	.002	(.015)
Male partner has higher education					.009	(.094)	.010	(.094)
Companionship
Intimacy							-.003	(-.035)
Control variables
Relationship duration	.001	(.001)	.001	(001)	.001	(.001)	.001	(.001)
Age of respondent and partner
Same age	ref		ref		ref		ref
Female respondent is older	.002	(.024)	.002	(.018)	.002	(.022)	.002	(.024)
Male partner is older	.001	(.011)	.001	(.006)	.001	(.003)	.001	(.002)
Birth cohort
1991-1993	ref		ref		ref		ref
1981-1983	.038	(.357)	.038	(.358)	.039	(.368)	.038	(.362)
1971-1973	-.008	(-.089)	-.008	(-.086)	-.006	(-.073)	-.007	(-.083)
N	3,272		3,272		3,272		3,272
Nagelkerke R2	.057		.058		.064		.064
Source: pairfam wave 1 2008/2009; calculations by the authors; average marginal effects (AME); logit coefficients in parenthesis; robust standard errors for interviewer ID; ***p<0.001; **p<0.01; *p<0.05; +p<0.1

 

Discussion and Conclusion

This paper extends the existing literature on the presence of an intimate partner during an interview. In contrast to most previous studies, the focus was on the causes of the partner’s presence and not on the potential influences on the response behaviour of the respondent. While previous studies on partner presence focused on structural reasons, we mainly investigated the association between characteristics of relationship dynamics and personality characteristics on the one hand and partner presence on the other hand. Descriptive findings revealed that in one of seven interviews the intimate partner was present. Furthermore, female partners were more often present than male partners.

Our multivariate results show that the opportunity structure was the most important determinant of partner presence during the interview (H1, H2). If the respondent or the partner was not employed, the probability of the partner’s presence increased. These results seem plausible for unemployed partners, but are surprising for unemployed respondents. Apparently, unemployed respondents did not choose specific time slots for the interview in which their partner was definitely not at home.

In the case of social control, the constellation of a male respondent and a female interviewer was associated with an increased presence of the female partner during the interview. One possible explanation is the control motivation of the female partner as a consequence of jealousy. Psychological studies have revealed that women are more affected by emotional jealousy than men, which might explain the absence of significant effects for men (Edlung et al. 2006). Furthermore, the respondent’s attractiveness was only associated with the presence of the female partner. However, the respondent’s attractiveness did not have the expected negative effect on female presence. Nevertheless, the mechanism of jealousy and attractiveness is unclear due to an important limitation of pairfam: there is no information on the attractiveness of the interviewer as a potential rival that would help to uncover the mechanism between attractiveness and jealousy (Buunk & Dijkstra 2004). Thus the hypothesis of social control (H3) was only partially confirmed.

In line with the social support theory, the presence of the female and the male partner was associated with the respondent’s self-worth (H5). One possible explanation for this association might be that if the self-worth is low, an interview might raise the respondent’s stress level. In this situation, the partner could act as a balance due to his/her support and could lower the respondent’s stress level. In this analysis, intimacy was only associated with the presence of the female partner, which partly confirmed Hypothesis 7.

A limitation of pairfam is that no information exists on whether the respondent initiated the presence of the partner or whether it was the partner himself/herself who initiated his/her presence. As Reuband (1987) showed, the partner himself/herself initiates his presence in only 22% of interviews. In all other cases, the presence of the partner is initiated by the respondent himself/herself, or jointly by the respondent and the partner, which should especially occur when the respondent is looking for social support. Furthermore, the paradata provided by pairfam do not include any information about the length of the partner’s presence during the interview. In addition, partner presence might bias the information on couple dynamics and partnership stability, so that the measurement error induced by the presence of the partner might explain the association between the explanatory variables used in this analysis and partner presence. But in pairfam all variables related to couple dynamics and partnership stability are obtained via CASI mode. Although CASI should minimize the bias induced by the partner’s presence, this is not fully guaranteed (Lavrakas 2008).

Factors (opportunity structure and social support) associated with the presence of the female partner might reflect traditional role models. Female partners in general might stay at home more often and also might be the most knowledgeable respondent in terms of specific partnership related questions, such as relationship duration, and be able to provide support during the interview.

In summary, the results illustrate that partner presence is mainly related to the opportunity structure. Intimacy and social support are also associated with partner presence, but it is not expected that these will induce a bias. As it has been reported in the literature that the partner’s presence may result in socially motivated misreporting (Aquilino 1993; Aquilino et al. 2000; Hartmann 1994; Tourangeau & Yan 2007), it might be easiest to prevent partner presence by taking into account the opportunity structure.

Interviewers could emphasize the importance to conduct the interview without a third person present when making the interview appointment. In pairfam, however, this is not the case because the chance that further respondents like the intimate partner participate in pairfam is higher when they are present during the interview. This benefits pairfam’s multi-actor design. But for surveys that collect data from only one respondent, it might be effective to prevent third person presence if the interviewer emphasizes the importance before the interview is conducted.

Analysing the reasons for partner presence in a cross-sectional design is a first step. Based on these results future research could investigate the development of the frequency of partner presence over subsequent waves. It would be interesting to analyse whether social control and social support matter in subsequent waves, since control motivation might decrease over time after the partner was present during the interview. The need for social support might decrease as well, as learning effects should occur if the interview situation has been repeated several times.

Appendix

Table A-1 Descriptive statistics of the independent variables

	Percent	Mean	SD
Living in the same household (1 = yes; 0 = no)	77.30
Respondent not employed (1 = yes; 0 = no)	19.34
Partner not employed (1 = yes; 0 = no)	13.56
Dominance of female partner (1-5)		3.21	.614
Dominance of male partner (1-5)		2.92	.650
Interview sex constellation (1 = opposite; 0 = same)	49.34
Attractiveness of the male respondent (1-7)		5.44	1.302
Attractiveness of the female respondent (1-7)		5.61	1.392
Dyadic coping of male partner (1-5)		4.21	.701
Dyadic coping of female partner (1-5)		4.28	.622
Self-worth of male respondent (1-5)		4.27	.661
Self-worth of female respondent (1-5)		4.08	.782
Educational level of couples
Woman higher educated	23.26
Man higher educated	30.71
Both have same education	46.03
Intimacy (1-5)		3.87	.774
Relationship duration (months)		95.48	72.11
Couple’s age
Woman older	17.05
Man older	72.74
Both have same age	10.21
Birth cohort
1991-1993	6.98
1981-1983	41.60
1971-1973	51.42
N	5620

 

Table A-2 Scales and operationalisation of the independent variables

Variable	Questionnaire	Operationalisation
Cohabitation	Cohabitation status	Both partners live in the same household vs. both partners live in separate households
Respondent non-working	Respondent´s employment status	Respondent non-working vs. respondent working
Partner non-working	Partner´s employment status	Partner non-working vs. partner working
Dominance (CASI)	How often does [name of current partner] get his/her way when you can’t agree on something? How often does [name of current partner] make you do things his/her way?	1: Never 5: Always
Interviewer-respondent sex constellation	Constellation of interviewer sex and respondent sex	Opposite sex vs. same sex
Attractiveness	How attractive do you find the respondent?	1: very unattractive 7: very attractive
Dyadic coping (CASI)	[name of partner] lets me know that he/she understands me. [name of partner] listens to me and gives me the chance to express myself. [name of partner] supports me in concrete ways when I have a problem.	1: Never 5: Always
Self-worth (CASI)	Sometimes I believe that I’m worthless. I like myself just the way I am. All in all, I am pleased with myself.	1: Not at all 5: Absolutely
Education	Respondent´s education and partner´s education (ISCED)	1:Respondent more educated 2: Partner more educated 3: Same education
Intimacy (CASI)	How often do you tell [name of current partner] what you’re thinking? How often do you share your secrets and private feelings with [name of current partner]?	1: Never 5: Always
Relationship duration		Months
Respondent´s age		Years
Partner´s age		Years
Birth cohort		1: 1991-1993 2: 1981-1983 3: 1971-1973

References

1. Aquilino, W. (1993): Effects of Spouse Presence During the Interview on Survey Responses Concerning Marriage. In: Public Opinion Quarterly 3 (57), P. 358–376.
2. Aquilino, W., Wright, D. & Supple, A. (2000): Response Effects Due to Bystander Presence in CASI and Paper-and-Pencil Surveys of Drug Use and Alcohol Use. In: Substance Use & Misuse 35 (6-8), P. 845–867.
3. Billiet, J. & Davidov, E. (2008): Testing the Stability of an Acquiescence Style Factor Behind Two Interrelated Substantive Variables in a Panel Design. In: Sociological Methods & Research 36 (4), P. 542–562.
4. Bodenmann, G. (2000). FDCT-N. Fragebogen zur Erfassung des dyadischen Copings als generelle Tendenz. Fribourg: Institut für Familienforschung und -beratung.
5. Buunk, B. & Dijkstra, P. (2004): Gender Differences in Rival Characteristics that Evoke Jealousy in Response to Emotional versus Sexual Infidelity. In: Personal Relationships 11(4), P. 395-408.
6. Cannell, C., Oksenberg, L. & Converse, J. (1977): Striving for Response Accuracy: Experiments in New Interviewing Techniques. In: Journal of Marketing Research 14 (3), P. 306–315.
7. Edlund, J.; Heider, J.; Scherer, C.; Farc, M.-M. & Sagarin, B. (2006): Sex Differences in Jealousy in Response to Actual Infidelity. In: Evolutionary Psychology 4 (1), P. 462-470.
8. Furman, W., & Buhrmester, D. (1985): Children’s Perceptions of the Personal Relationships in their Social Networks. In: Developmental Psychology 21 (6), P. 1016–1024.
9. Guerrero, L. K., Spitzberg, B. H., & Yoshimura, S. M. (2004). Sexual and Emotional Jealousy. In Harvey,J.; Wenzel, A. & Sprecher, S. (Eds.), The Handbook of Sexuality in Close Relationships. Mahwah, NJ:
10. Lawrence Erlbaum Associates, P. 311-345
11. Hartmann, P. (1994): Interviewing when the Spouse is Present. In: International Journal of Public Opinion Research 6 (3), P. 298–306.
12. Huinink, J., Brüderl, J., Nauck, B., Walper, S., Castiglioni, L. & Feldhaus, M. (2011): Panel Analysis of Intimate Relationships and Family Dynamics (pairfam): Conceptual Framework and Design. In: Zeitschrift für Familienforschung 23 (1), P. 77–101.
13. Johnson, T. (2014): Sources of Error in Substance Use Prevalence Surveys. In: International Scholarly Research Notices.
14. Krumpal, I. (2013): Determinants of Social Desirability Bias in Sensitive Surveys: a Literature Review. In: Quality & Quantity 47 (4), P. 2025–2047.
15. Lakey, B. & Cohen, S. (2000): Social Support Theory and Measurement. In: Cohen, S., Brittney, L. & Gottlieb, B. (Eds.): Social Support Measurement and Intervention. A Guide for Health and Social Scientists. Oxford, New York: Oxford University Press, P. 29–52.
16. Lander, B. (2000): Anwesenheitseffekte im Wandel. Eine Sekundäranalyse zur Anwesenheit des Partners im Interview anhand des ALLBUS 1980 bis 1998. In: Zeitschrift für Soziologie 29 (3), P. 227–238.
17. Lavrakas, P. (2008): Encyclopedia of Survey Research Methods. Thousand Oaks: SAGE (Volume 1).
18. Long, J. & Freese, J. (2014): Regression Models for Categorical Dependent Variables using Stata. 3. ed. College Station, Tex.: Stata Press.
19. Mohr, H. (1986): Dritte beim Interview. Ergebnisse zu Indikatoren aus dem Bereich Ehe und Partnerschaft mit den Daten des “Wohlfahrtssurvey 1984”. In: ZA Information 19, P. 52–71.
20. Mood, C. (2010): Logistic Regression. Why We Cannot Do What We Think We Can Do, and What We Can Do About It. In: European Sociological Review 26 (1), P. 67–82.
21. Reuband, K-H. (1984): Dritte Personen beim Interview – Zuhörer, Adressaten oder Katalysatoren der Kommunikation? In: Meulemann, H. & Reuband, K-H. (Ed.): Soziale Realität im Interview. Frankfurt: Campus, P. 117-156.
22. Reuband, K.-H. (1987): Unerwünschte Dritte beim Interview. In: Zeitschrift für Soziologie 16 (4), P. 303-308.
23. Reuband, K.-H. (1992): In Third Persons in the Interview Situation and their Impact on Responses. In: International Journal of Public Opinion Research 4 (3), P. 269-274.
24. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton: Princeton University Press.
25. Schneider, S. (2008): Applying the ISCED-97 to the German educational qualifications. In: Schneider, S. (Ed.): The International Standard Classification of Education (ISCED-97). An Evaluation of Content and Criterion Validity for 15 European Countries. Mannheim: MZES, P. 76–102.
26. Thompson, S. (1981): Will it Hurt Less if I Can Control it? A Complex Answer to a Simple Question. In: Psychological Bulletin 90 (1), P. 89–101.
27. Tourangeau, R. & Yan, T. (2007): Sensitive Questions in Surveys. In: Psychological Bulletin 133 (5), P. 859–883.