Abstract
Measurements of sexual intercourse frequency are informative for research on pregnancy, contraception, and the transmission of sexually transmitted infections; however, efficiently collecting data on this sensitive topic is complex. The purpose of this study was to determine whether retrospective recall of sexual intercourse frequency was consistent with information obtained through the use of prospective daily diary methods corresponding to the same time period in a diverse sample of women. A total of 185 women participated in a longitudinal, prospective cohort study of oral contraceptive users and 98 of these women provided complete information on sexual intercourse frequency on diaries (prospective) and postcards (retrospective). Linear mixed models were used to test for variation in response within categories of demographic and other variables. The mean number of days women had sexual intercourse per week was 1.5 days using prospective diary information versus 2.0 days when using 3-month retrospective recall (p <0.001). Mean differences for the various sociodemographic subgroups were positive for all groups indicating that women consistently reported a higher frequency of sexual intercourse on the retrospective postcards than they recorded on their prospective diaries; however, these mean differences did not vary significantly. If confirmed in other samples, the use of retrospective methods may be adequate to accurately collect data on sexual intercourse frequency—and may be preferable. Using only retrospective measurements could decrease study costs, the burden to participants, and have a higher response rate.
Keywords: Measurement of sexual intercourse, Sexual behavior, Coitus, Reproducibility of results
Introduction
Since the 1980s, measurements of sexual intercourse frequency have been acknowledged as informative for research on pregnancy, contraception, and the transmission of sexually transmitted infections [1–4]. Efficiently collecting data on this sensitive topic is complex. Numerous studies have compared the reliability of reporting vaginal intercourse frequency using self-reported, prospective daily diary methods versus retrospective recall [1–9]. Results have been inconsistent, with some studies finding moderately strong agreement between diary and recall methods [3, 4] especially among older women [10] and with the use of simpler diary layouts [8]; some studies suggesting moderate associations [5]; and others suggesting poor agreement between the two methods [1, 2, 7]. Table 1 provides a comparison of selected studies on the measurement of sexual intercourse frequency that were conducted in adult, heterosexual populations.
Table 1.
Comparison of select studies on the measurement of sexual intercourse frequency conducted in adult, heterosexual populations
| Study | Study participants | Diary methodology | Main findings |
|---|---|---|---|
| Hornsby and Wilcox [1] | N = 91, women aged 21–38; primarily NHWa | Daily for 4–12 weeks; compared vaginal intercourse reported at baseline interview (average times having intercourse/week) versus prospective daily record cards (yes/no); measurements did not coincide with same time period | Reported significantly higher frequency at baseline than on the daily record cards |
| Leigh et al. [3] | N = 79, men and women aged 16–38; primarily NHW | Daily for 4 weeks; compared vaginal intercourse reported on 3 types of daily dairies (2 phone methods, 1 written diary; yes/no to sex and condom use) versus recall 1 month later | Diary reports and retrospective recall were strongly correlated |
| Hays et al. [8] | N = 89, women aged 18–25; primarily NHW | Daily for 6 weeks; compared vaginal intercourse reported on 3 types of daily dairies (all written at various levels of complexity; yes/no to sex and several additional questions such as condom use) and phone regimen versus recall 6–8 weeks later | Diaries with simpler layouts had high concordance with retrospective recall |
| Steiner et al. [9] | N = 195; women aged 18–40; racial/ethnic distribution unknown | Daily for 4 weeks; compared vaginal intercourse reported at baseline interview (average frequency of intercourse during a month) to prospective data collected during the study (yes/no and condom use); measurements did occur over same time period | No significant difference in frequency reported at baseline and in daily diaries |
| Graham et al. [7] | N = 75, men and women aged 19–35; primarily NHW | Daily for 4 weeks; compared sexual activity (i.e. oral, vaginal, etc.) and condom use reported in the daily dairy to recall at 1, 2, or 3 months later | Recall error for vaginal intercourse was significantly greater at 3 months than at 1 month post-diary |
NHW non-hispanic white
Many prior studies have had features that may limit the generalizability of their results. Examples include limited racial diversity [1, 3, 7, 8, 11]; use of sex workers [4]; use of sexual behavior among homosexuals [2]; and lack of randomized control study design [1, 2, 4, 7, 9–11]. Furthermore, there may be methodological issues in some studies regarding the actual retrospective and prospective measurements used to assess sexual intercourse frequency. Specifically, some studies have used predictions of sexual activity instead of actual retrospective recall [6] or have compared retrospective and prospective measurements that do not coincide with the same time period [1, 9].
Also, many studies have focused on adolescents or young adults [5–7], with less research on adults. Lastly and importantly, the majority of the research to date on the validity of retrospective versus prospective methods of sexual activity recall was conducted over 10 years ago [1–5, 8–10]. Since that time there have been considerable changes in health behaviors and attitudes towards contraception and sexual intercourse. For example, women in the United States have gained increased access to newer forms of contraception including intrauterine devices, contraceptive rings, and implants [12]. Continuing shifts in gender roles and egalitarianism in marriages also have been associated with changes in frequencies of sexual intercourse [13]. Thus, potentially informative contemporary research on sexual intercourse frequency is lacking.
The current study aimed to determine whether retrospective recall of vaginal sexual intercourse frequency was consistent with information obtained through the use of prospective daily diary methods corresponding to the same relevant time period in an adult sample of women aged 18–40 who were diverse with respect to race/ethnicity and socioeconomic status. Given the noted complexities and costs associated with collecting sexual intercourse frequency data, as well as the limited generalizability and potentially outdated information of existing studies, this study aimed to provide additional methodological insight on how to best obtain information on this sensitive topic.
Methods
Study Design and Population
The Fertility and Oral Contraceptive Use Study was a feasibility study designed to explore methodological issues related to the future development and implementation of a large, prospective cohort study to investigate the role of obesity in oral contraceptive failure. The feasibility work included evaluation of issues related to the collection and measurement of sensitive data such as frequency of sexual intercourse. The study was approved by the local institutional review board.
From June 2009 to January 2011, women were recruited from waiting rooms of three sites: a family medicine clinic, an obstetrics and gynecology clinic, and a family planning clinic. Additionally, fliers were posted in the clinics and informational letters were sent to a subset of women who had been seen at the clinics for routine care during the 2009 calendar year. The family medicine and obstetrics and gynecology clinics are affiliated with a local hospital and the family planning clinic is associated with a local health department. All of the clinics serve a large, urban city in North Carolina.
The hospital system is the largest healthcare system in the Carolinas, and one of the largest public, non-profit healthcare systems in the nation. The obstetrics and gynecology clinic affiliated with this hospital has over 34,000 patient visits each year. Approximately 25 % of the patients seen at this clinic are of Hispanic or Latina ethnicity. In terms of racial categories, the majority of women are African American or black (37 %) or White (61 %). An estimated 40 % of patients are self-pay and 44 % use Medicaid or Medicare. The family medicine clinic affiliated with this hospital system has approximately 25,000 active patients. An estimated 5 % of patients are of Hispanic or Latina ethnicity and in terms of racial categories, the clinic has the following distribution: 65 % African American or black, 28 % White, 5 % Asian, and 2 % American Indian or Alaska native. Thirty percent of patients are self-pay and 40 %use Medicaid or Medicare. The family planning clinic affiliated with the local health department serves a large, urban area where the majority of residents are Non-Hispanic white. Approximately 15 %of residents are Non-Hispanic black and 6 % are Hispanic. This clinic has over 3,000 patient visits each year. The racial/ethnic distributions of these three clinics are similar to the distributions seen not only for the state of North Carolina, but also for other states in the Southeast region [14].
Women were eligible to participate in this feasibility study if they met the following criteria: English- or Spanish- speaking; between the ages of 18–40 years; currently using oral contraceptives; not planning to move out of the area in the next year; never had a sterilization procedure (woman or her partner); and had never been told by a doctor that they were infertile (woman or her partner).
Recruitment and Data Collection
A total of 185 women provided written informed consent and participated in the study. The baseline interview was conducted by trained, bilingual research assistants on site at the clinics. It took approximately 15–20 minutes to complete and included questions relating to the participants’ sexual activity, contraceptive use, and demographic information. In addition, the research assistants obtained anthropometric measurements from participants using standardized procedures and equipment. Women were then given a set of diaries with detailed instructions and stamped envelopes to mail them back to the researchers. The diary was one large sheet of paper and included a simple chart for women to check vaginal sexual intercourse and oral contraceptive use for each day. Participants were given three diaries, one for each month. Women were asked to complete the diaries on a daily basis and mail one back each month as it was completed.
Additional follow-up points for this feasibility study occurred at 3-month intervals out to 15 months. For this analysis, we used data from the 3-month follow-up. The 3-month follow-up was a postcard that asked about oral contraceptive use, adherence, and sexual activity during the prior 3 months. The cards were simple and included only the three questions and used a convenient, confidential method to provide the information. Participants checked or wrote their answers, ripped off the pre-stamped postcard, and placed it in the mail. The return postcards did not have participants’ names (only a study ID so that the information could later be linked to the participants’ interview and questionnaire data) and were designed so that the information mailed back did not include the original questions–they merely contained checkmarks by set responses (e.g. yes or no) or an actual numeric variable (e.g. a woman may have written ‘‘4’’ to indicate that she had had sexual intercourse 4 days in a week).
In order to maximize response rates, trained interviewers called participants who did not return their diaries and/or 3-month postcard to remind them to return their study materials or to see if they needed a replacement diary or postcard. Participants were compensated for their participation with gift cards to a local supermarket.
Description of Variables
During the baseline interview, women self-reported information on the following variables: age, marital status, education, race/ethnicity (derived from two separate questions on ethnicity [i.e. Hispanic or Latina or not Hispanic or Latina] and race [i.e. Black or African American, White, Asian, American Indian or Alaska Native, or Native Hawaiian or Pacific Islander] and ultimately grouped into Hispanic, Non-Hispanic white, Non-Hispanic black, and other), smoking, alcohol consumption, number of people residing in household, income, and parity. These variables were evaluated as covariates in the analysis as they were potential confounders in prior studies that have assessed agreement between prospective and retrospective measures of sexual intercourse [1, 9, 10].
To assess the validity of retrospective recall of vaginal sexual intercourse frequency, the diary card information was compared to the frequency of sexual intercourse information provided on the 3-month postcard employing analytic approaches previously used by Hornsby and Wilcox in their study on the validity of sexual intercourse [1]. On the current diary cards, women checked if they had vaginal sexual intercourse on that particular day. The 3-month postcard asked women the following: ‘‘During the last 3 months, how many days did you have sexual intercourse resulting in vaginal penetration in a typical week?’’ The frequencies given on the prospective diary cards were averaged to obtain an average weekly frequency of sexual intercourse. This weekly average was then subtracted from the weekly frequency reported on the retrospective 3-month postcard to produce a difference value (i.e. post-card value-diary value). We hypothesized that women would overestimate their sexual frequency on the retrospective postcards resulting in positive difference values.
Statistical Analysis
Although a total of 108 (58.4 %) women completed all 3 diaries and the 3-month follow-up survey, 10 women incorrectly interpreted the intercourse frequency question on the follow-up survey and reported having sex more than 7 days per week. Thus, after excluding these women, 98 women remained for the analysis. Frequencies and percentages were used to describe the baseline characteristics of the study population. A paired t test was used to test whether the difference values obtained by subtracting the diary weekly average (prospective) from the weekly frequency reported on the 3-month postcard (retrospective) were 0 on average. Finally, linear mixed models (to adjust for correlated outcome measures) were used to test for variation in response within categories of demographic and other variables.
Results
The majority of the study population was between 25 and 34 years old (51.0 %) and single (59.2 %; Table 2). Based on diary information, the mean number of days women had sexual intercourse per week was 1.5 days. Based on 3-month retrospective recall, the mean number of days of sexual intercourse per week was marginally higher (mean: 2.0 days). This mean difference was statistically significant (p <0.001). With respect to mean differences for various sociodemographic groups, the mean differences were positive for all groups indicating that women consistently reported a higher frequency of sexual intercourse on retrospective postcards than they recorded on their prospective diaries (Table 2). The mean differences did not vary significantly for the various sociodemographic groups examined. Approximately 11 % (n = 11) of women had perfect agreement between their postcard and diary information while nearly 62 % (n = 61) of women’s information agreed within ± 1.0 day (Table 3). When agreement was considered to be within ± 2.0 days of each other, almost 85 % (n = 83) of women’s diary and retrospective recall information were in agreement.
Table 2.
Mean weekly frequency of sexual intercourse from 3-month postcard (retrospective) and diary (prospective) information; Fertility and Oral Contraceptive Use Study (FOCUS)
| Group | Sample size | Percenta | Postcard (retrospective) |
Diary (prospective) |
Differencea (postcard-diary) |
95 % CI for difference |
|---|---|---|---|---|---|---|
| Overall | 98 | – | 2.0 | 1.5 | 0.6 | 0.3, 0.9 |
| Age | ||||||
| <25 | 34 | 34.7 | 1.7 | 1.4 | 0.3 | −0.4, 1.0 |
| 25–34 | 50 | 51.0 | 2.2 | 1.6 | 0.6 | 0.4, 0.9 |
| >35 | 14 | 14.3 | 2.1 | 1.1 | 1.1 | 0.0, 2.1 |
| Marital status | ||||||
| Married | 21 | 21.4 | 2.4 | 1.6 | 0.9 | 0.4, 1.4 |
| Living with partner | 19 | 19.4 | 2.9 | 2.6 | 0.3 | −0.3, 1.0 |
| Single | 58 | 59.2 | 1.6 | 1.1 | 0.6 | 0.1, 1.0 |
| Educational level | ||||||
| High school or less | 31 | 31.6 | 2.7 | 2.0 | 0.7 | 0.1, 1.2 |
| Some college | 34 | 34.7 | 1.6 | 1.4 | 0.2 | −0.3, 0.7 |
| College graduate or more | 33 | 33.7 | 1.9 | 1.0 | 0.9 | 0.5, 1.3 |
| Income | ||||||
| <$30,000 | 55 | 56.1 | 2.1 | 1.5 | 0.6 | 0.2, 1.0 |
| $30,001–$45,000 | 18 | 18.4 | 2.3 | 1.7 | 0.6 | −0.3, 1.5 |
| >$45,000 | 25 | 25.5 | 1.7 | 1.3 | 0.5 | 0.0. 0.9 |
| Number in household | ||||||
| 1 person | 11 | 11.2 | 1.5 | 0.9 | 0.5 | −0.5, 1.5 |
| 2 people | 34 | 34.7 | 1.7 | 1.3 | 0.4 | 0.0. 0.8 |
| >3 people | 53 | 54.1 | 2.4 | 1.7 | 0.7 | 0.2, 1.2 |
| Race/ethnicity | ||||||
| Hispanic | 12 | 12.2 | 2.4 | 2.2 | 0.2 | −0.4, 0.9 |
| Non-Hispanic white | 36 | 36.7 | 1.9 | 1.3 | 0.6 | 0.2, 1.0 |
| Non-Hispanic black | 44 | 44.9 | 2.1 | 1.4 | 0.7 | 0.1, 1.2 |
| Other | 6 | 6.1 | 1.8 | 1.3 | 0.6 | −0.7, 1.9 |
| Alcohol consumption | ||||||
| Yes | 46 | 46.9 | 2.1 | 1.5 | 0.6 | 0.2, 1.1 |
| No | 52 | 53.1 | 2.0 | 1.5 | 0.5 | 0.1, 0.9 |
| Smoking | ||||||
| Yes | 17 | 17.4 | 2.8 | 1.9 | 0.9 | 0.1, 1.6 |
| No | 81 | 82.7 | 1.9 | 1.4 | 0.5 | 0.2, 0.8 |
| Parity | ||||||
| 0 | 44 | 44.9 | 1.9 | 1.3 | 0.7 | 0.2, 1.1 |
| 1 | 24 | 24.5 | 1.9 | 1.4 | 0.4 | 0.0, 0.8 |
| ≥2 | 30 | 30.6 | 2.4 | 1.8 | 0.6 | 0.0, 1.2 |
Values rounded
Table 3.
Distribution of difference values between retrospective (postcard) and prospective (diary) information on weekly frequency of sexual intercourse; Fertility and Oral Contraceptive Use Study (FOCUS)
| Retrospective (postcard)— prospective (diary) |
Number | Percent |
|---|---|---|
| ≤2.0 | 4 | 4.1 |
| −2.0 to −1.1 | 6 | 6.1 |
| −1.0 to −0.6 | 2 | 2.0 |
| −0.5 to −0.1 | 13 | 13.3 |
| 0 | 11 | 11.2 |
| 0.1 to 0.5 | 17 | 17.3 |
| 0.6 to 1.0 | 18 | 18.4 |
| 1.1 to 2.0 | 16 | 16.3 |
| >2.0 | 11 | 11.2 |
Discussion
Women reported slightly more days of sexual intercourse per week on the retrospective 3-month postcards than on the prospective daily dairies. Mean differences were positive for all sociodemographic groups and did not vary significantly. While only 11 % of women had perfect agreement between their retrospective and prospective weekly frequency of sexual intercourse information, this agreement increased to nearly 85 % when a ± 2.0 day window was considered.
The majority of previous studies examining the validity of sexual recall in adult populations have found that prospective daily recall and retrospective recall of sexual frequency are in agreement (Table 1) [3, 4, 8, 9]. Similar to other studies conducted among women of reproductive age, we found that participants tended to report a slightly higher frequency of sexual intercourse retrospectively as compared to prospectively [1, 9]. However, it is important to note that unlike our study, these previous studies did not use retrospective and prospective measurements that coincided with the same time period. Consistent with our findings, other studies have also noted that differences in retrospective and prospective measurements of sexual intercourse frequency do not vary significantly for various demographic groups [1] and that diaries with the simplest layouts have the best agreement with retrospective recall of sexual intercourse frequency [8].
The current study has several strengths including the use of a racially/ethnically and socioeconomically diverse sample of adult women. Many prior studies focused on adolescents [5–8]. Numerous studies also lacked racial and ethnic diversity or had higher socioeconomic status samples [1, 3, 7–9], and it is important to understand how retrospective and prospective information on frequency of sexual intercourse compares in a diverse sample of adult women in order to improve generalizability. Another strength of this study is the fact that the retrospective and prospective measurements were for the same time period. Other studies have assessed the validity of retrospective recall of sexual intercourse by considering time periods that do not coincide with one another [1, 9].
There are some limitations to this study, as well. Misclassification is possible since women self-reported their frequency of sexual intercourse on both prospective and retrospective survey items. Women were to complete their diaries daily at home and mail them to the study staff upon completion. Women may not have actually completed the diaries on a prospective, daily basis. For example, a woman may have filled in her diaries at the end of each week instead of daily. It is also possible that women may have incorrectly reported their sexual frequency since this is a sensitive topic. Furthermore, there were some problems with comprehending the question on the postcard. Ten women had to be excluded from the analysis due to the fact that they reported having sexual intercourse more than 7 days per week. Presumably these women misread the question and answered how many times they had sexual intercourse during the week. We investigated whether these women differed from other participants and found no differences with respect to a number of sociodemographic characteristics.
An additional concern, one that potentially would affect the internal validity of a study using this variable as an exposure of interest as well as the decision about which data collection approach to use, is that not all women enrolled in the feasibility study completed all the follow-up questionnaires necessary for this particular analysis. The overall response rates for the diaries in the feasibility study were fairly high, but they did decline by the third diary, perhaps due to fatigue associated with completing daily diaries (Diary 1: 77.8 %, Diary 2: 74.1 %, and Diary 3: 63.4 %); presumably, for a longer study the response rate would decline further. In contrast, the overall response rate for the 3-month postcard in the feasibility study was nearly 80 %. Women who completed all aspects of the study may be different from women who chose not to respond to subsequent follow-up surveys. We compared study participants who completed all the diaries and 3-month follow-up postcard to women who did not and found no statistically significant differences between the two groups with respect to various sociodemographic factors including race/ethnicity, age, parity, income, marital status, and education.
Despite these limitations, these findings suggest that retrospective recall of sexual intercourse frequency may be an appropriate method to use to collect data on this sensitive topic. Retrospective recall may benefit not only study participants, but also investigators. The participant burden associated with completing a daily diary for 3 months is considerable. As seen in our study, response rates associated with the diaries dropped each month of data collection. Other investigators have noted that participant fatigue in completing diaries may impact the quality of data since participants may become less thorough in their reporting over time [15, 16].
In addition to concerns about data quality, study costs are also a concern for investigators. In this feasibility study of 185 women, we spent nearly $11,500 on costs associated with collecting the prospective diary information (i.e. printing, mailing, and participant incentive costs). In comparison, $4,100 was spent on the same associated costs in terms of collecting the retrospective postcard information. Neither estimate accounts for the costs associated with hiring research staff to recontact and/or resend study materials to participants who did not return study materials, however, these costs would be higher for the diaries since they require more mailings due to the fact they are returned on a monthly basis.
As previously indicated, this feasibility study sought to investigate methodological issues that would be associated with implementing a larger, prospective cohort study to investigate the possible obesity-oral contraceptive failure association in the future. Based on our findings, it may be reasonable to include a one-item question on sexual intercourse frequency on a baseline interview and/or several retrospective questions throughout the course of a longitudinal study, which would not only further reduce participant burden and study costs, but also possibly improve the quality of data since there would not be less concerns related to non-response or loss to follow-up. It is important to note, however, that retrospective recall may not be appropriate for all areas of research. While we found that 85 % of women’s information on sexual intercourse frequency was in agreement when a ±2.0 day window was considered, this window may not be appropriate for all areas of research. In studies of fertility, a larger window may be acceptable due to the larger number of days associated with the ‘‘fertile window.’’ However, for studies of sexually transmitted infections a smaller window may be necessary to better assess risk.
In summary, there were some differences between prospective diary information and retrospective recall of weekly frequency of sexual intercourse. Women tended to report a higher weekly frequency of sexual intercourse on retrospective postcards than they recorded on daily diaries, and this was consistent across numerous sociodemographic variables. The great majority of women’s information on intercourse frequency was in agreement when a ±2.0 day window was considered. Similar to the findings of Hornsby and Wilcox, this suggests that any resulting bias is likely to be nondifferential [1].
Future studies may consider using alternative methods to paper diaries given the fatigue associated with completing daily dairies. For example, the use of text messaging may be an option to record sexual intercourse information prospectively. If diaries are preferable, investigators may consider monetary incentives that include a ‘‘bonus’’ if a participant returns all of her diaries. While the focus of this study was on oral contraceptive users, other studies may also enroll women using other methods of contraception, or no method at all, to determine if sexual intercourse frequency and/or agreement between prospective and retrospective measures of sexual intercourse frequency differ by contraceptive method choice. If confirmed in other large, racially/ethnically diverse samples, the use of retrospective methods at selected intervals may be adequate to accurately collect data on frequency of sexual intercourse as part of prospective research—and may be preferable. Using only retrospective measurements could decrease study costs, the burden to participants, and have a higher response rate.
Acknowledgments
This work was supported by the National Institutes of Health Grant 1R21HD056173-01A2. We thank the Fertility and Oral Contraceptive Use Study (FOCUS) research staff for their assistance with this project.
Contributor Information
Larissa R. Brunner Huber, Email: lrhuber@uncc.edu, Department of Public Health Sciences, UNC Charlotte, Charlotte, NC, USA.
Jordan E. Lyerly, Department of Public Health Sciences, UNC Charlotte, Charlotte, NC, USA
Ashley M. Young, Department of Public Health Sciences, UNC Charlotte, Charlotte, NC, USA
Jacek Dmochowski, Department of Mathematics and Statistics, UNC Charlotte, Charlotte, NC, USA.
Tara M. Vick, Department of Obstetrics and Gynecology, Carolinas Medical Center, Charlotte, NC, USA
Delia Scholes, Group Health Research Institute, Group Health Cooperative, Seattle, WA, USA.
References
- 1.Hornsby PP, Wilcox AJ. Validity of questionnaire information on frequency of coitus. American Journal of Epidemiology. 1989;130:94–99. doi: 10.1093/oxfordjournals.aje.a115326. [DOI] [PubMed] [Google Scholar]
- 2.Coxon AP. Parallel accounts? Discrepancies between self-report (diary) and recall (questionnaire) measures of the same sexual behaviour. AIDS Care. 1999;11:221–234. doi: 10.1080/09540129948108. [DOI] [PubMed] [Google Scholar]
- 3.Leigh BC, Gillmore MR, Morrison DM. Comparison of diary and retrospective measures for recording alcohol consumption and sexual activity. Journal of Clinical Epidemiology. 1998;51:119–127. doi: 10.1016/s0895-4356(97)00262-x. [DOI] [PubMed] [Google Scholar]
- 4.Ramjee G, Weber AE, Morar NS. Recording sexual behavior: Comparison of recall questionnaires with a coital diary. Sexually Transmitted Diseases. 1999;26:374–380. doi: 10.1097/00007435-199908000-00002. [DOI] [PubMed] [Google Scholar]
- 5.Minnis AM, Padian NS. Reliability of adolescents’ self-reported sexual behavior: A comparison of two diary methodologies. Journal of Adolescent Health. 2001;28:394–403. doi: 10.1016/s1054-139x(00)00218-4. [DOI] [PubMed] [Google Scholar]
- 6.Shrier LA, Shih MC, Beardslee WR. Affect and sexual behavior in adolescents: A review of the literature and comparison of momentary sampling with diary and retrospective self-report methods of measurement. Pediatrics. 2005;115:e573–e581. doi: 10.1542/peds.2004-2073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Graham CA, Catania JA, Brand R, Duong T, Canchola JA. Recalling sexual behavior: A methodological analysis of memory recall bias via interview using the diary as the gold standard. Journal of Sex Research. 2003;40:325–332. doi: 10.1080/00224490209552198. [DOI] [PubMed] [Google Scholar]
- 8.Hays MA, Irsula B, McMullen SL, Feldblum PJ. A comparison of three daily coital diary designs and a phone-in regimen. Contraception. 2001;63:159–166. doi: 10.1016/s0010-7824(01)00183-4. [DOI] [PubMed] [Google Scholar]
- 9.Steiner MJ, Hertz-Picciotto I, Taylor D, Schoenbach V, Wheeless A. Retrospective vs. prospective coital frequency and menstrual cycle length in a contraceptive effectiveness trial. Annals of Epidemiology. 2001;11:428–433. doi: 10.1016/s1047-2797(01)00248-4. [DOI] [PubMed] [Google Scholar]
- 10.Lauritsen JL, Swicegood CG. The consistency of self-reported initiation of sexual activity. Family Planning Perspectives. 1997;29:215–221. [PubMed] [Google Scholar]
- 11.Anderson BL, Broffitt B. Is there a reliable and valid self-report measure of sexual behavior? Archives of Sexual Behavior. 1988;17:509–525. doi: 10.1007/BF01542339. [DOI] [PubMed] [Google Scholar]
- 12.Jones J, Mosher W, Daniels K. Current contraceptive use in the United States 2006–2010, and changes in patterns of use since 1995. National Health Statistics Reports. 2012;60:1–25. [PubMed] [Google Scholar]
- 13.Kornich S, Brines J, Leupp K. Egalitarianism, housework, and sexual frequency in marriage. American Sociological Review. 2012;78:26–50. doi: 10.1177/0003122412472340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kaiser Family Foundation. [Accessed 5 Sept 2013];State health facts: Population distribution by race/ethnicity. http://kff.org/other/state-indicator/distribution-by-raceethnicity/#table.
- 15.Verbrugge LM. Health diaries. Medical Care. 1980;18:73–95. doi: 10.1097/00005650-198001000-00006. [DOI] [PubMed] [Google Scholar]
- 16.Wiseman V, Conteh L, Matovui F. Using diaries to collect data in resource-poor settings: Questions on design and implementation. Health Policy Plan. 2005;20:394–404. doi: 10.1093/heapol/czi042. [DOI] [PubMed] [Google Scholar]