Sex research - a summary of chapter 3 of Understanding human sexuality by Hyde and DeLamater

Sexology
Chapter 3
Sex research

There are different types of sex research, but basically the techniques vary in terms of the following:

• How sexuality is measured
  • Self-reports
  • Observations of behaviour
  • Using biological measures
• Whether large numbers of people are studies in surveys, or whether a smaller number of people are studies
• Whether the studies are conducted in the laboratory or in the field
• Whether sexual behaviour is studied as it occurs naturally or whether some attempt is made to manipulate it in an experiment.

Measuring sex

The first thing that researchers have to decide is how the measure the particular aspect of sexuality they want to study.

Self-reports

The most common method for measuring sexuality is self-reports, in which the participants are asked questions about their sexual behaviour.

Self-reports can be collected in a number of ways

• Paper questionnaires
• Interviews
• Online

Behavioural measures

Several alternatives are available for behavioural measures of sexuality.

• Direct observation
  A behavioural measure in which the scientist directly observes the behaviour being studied
• Eye-tracking
  A behavioural measure in which a device measures the participant’s point of gaze over time
• Police reports
  • The problem with this measure is that it detects only those cases that are reported by the police

Implicit measures

Implicit Association Test (IAT)
Measures an individual’s relative strength of association between different pair of concepts with reaction times.
We react quicker to two concepts that we associate strongly.
People cannot fake reaction times.

Biological measures

Genital measures: assess arousal by using devices that measure erection in males and vaginal changes in females.

• In males, penile plethysmography is used
• In females, a vaginal photoplethysomgraph is used

MRI and fMRI are being used increasingly in sex research.
Two limitations to use of fMRI in sex research

• Participants must be lying very still to get good images
• fMRI depends on contrasting the difference between two stimuli, therefore the results depend very much on the choice of each set of stimuli

Measures of pupil dilation.
Our pupils dilate when we look at something that is interesting or arousing, or that puts a big load on our brain.

Issues in sex research

Sampling

An important step in conducting sex research is to identify the appropriate population of people studied.
Generally the scientist is unable to get data for all people in the population so a sample is taken.

If a sample is a random sample or representative of the population in question and if it is a reasonably large sample, then results obtained can safely be generalized to the population that was originally identified.

Probability sampling: an excellent method of sampling in which each member of a population has a known probability of being included in the sample.

Typically sampling proceeds of three phases

• The question is identified
• A method for obtaining a sample is adopted
• The people in the sample are contacted and asked to participate

Problem of refusal or nonresponse: the problem that people will refuse to participate in a sex survey, this making it difficult to study a random sample.
Volunteer bias: a bias in the results that arises when some people refuse to participate, so that those who are in the sample are volunteers who may in some way differ from those who refuse to participate

Convenience sample: a sample chosen in a haphazard manner relative to the population of interest. Not a random op probability sample.

Accuracy of measurement

Purposeful distortion

Purposeful distortion: purposely giving false information in a survey.
These distortion can be in two directions

• Enlargement
  Exaggerating sexual activity
• Concealment
  Minimizing or hiding the fact that people have don certain things

Distortion is a basic problem when using self-reports
To minimize distortion, participants must be impressed with the fact that because the study will be used for scientific purposes, their reports must be as accurate as possible.
They must also be assured that their responses will be completely anonymous.

Memory

Some of the questions asked in sex surveys require respondents to recall what their sexual behaviour was like many years before.
It might be difficult to remember things correctly.

A way to avoid memory problems is the use of daily diaries, in which people report behaviour each day.

Difficulties with estimates

In some sex surveys people are asked to give estimates of things that they probably cannot estimate very accurately.

Evidence on reliability of self-reports

• Test-retest reliability.
• Inter-rater reliability
• Computer-assisted self-interview (CASI)
  A method of data collecting in which the respondent fills out questionnaires on a computer. Headphones and a soundtrack reading the questions can be added for young children or poor readers.

Accuracy of behavioural observations

Self-reports may be inaccurate, although the evidence indicates that they are generally accurate.
Direct observations are accurate. No purposeful distortion or inaccurate memory can intervene.

Problems of direct observation

• Expensive and time-consuming
  • Because of this only a small sample is measured
• Obtaining a random sample or probability sample is difficult

Extraneous factors

Various extraneous factors may influence the outcomes of sex research

• Gender, race or age of the interviewer
• Simple factors as the wording of questions

Ethical issues

Ethical problems are particularly difficult in sex research, because people are more likely to feel that their privacy has been invaded when you ask them about sex.

The ethical standards in U.S involve three basic principles

• Informed consent
  People have a right to be informed, before participating, of what they will be asked to do in the research
  They may not be forced to participate or be forced to continue.
• Protection from harm
  Investigators should minimize the amount of physical and psychological stress to people in their research.
  He principle of anonymity of response is important to ensure that participants will not suffer afterwards for their participation in research
• Justice
  Justice principle: the risks of participation should be distributed fairly across groups in society, as should the benefits.

A cost-benefit approach

Weighing the costs of the research against the benefits of the research.

The major sex surveys

The Kinsey report

The sample

Kinsey and his colleagues interviewed a total of 5,300 males and 5,940 females.
Only interviews with whites were included in the publications.

Initially, Kinsey was not concerned with sampling issues, his goal was to collect sex histories from as wide a variety of people as possible.

The interviews

Kinsey’s face-to-face interviewing techniques are highly regarded.
The interviewing techniques were probably very successful in minimizing purposeful distortion.
Kinsey took strict precautions to ensure that respondents were anonymous and remained anonymous.

How accurate were the Kinsey statistics?

It is impossible to say how accurate the Kinsey statistics are.

The NHSLS

The National Health and Social Life Survey, NHSLS.
The research involved as probability sample of households in the U.S.

The data were obtained in face-to-face interviews supplemented by brief written questionnaires.

The NHSLS is one of the best sex surveys of the general population of the U.S that we have today.

The NSSHB

National Survey of Sexual Health and Behaviour (NSSHB)
Most recent major national U.S sex survey.

Methods of probability sampling were excellent.
50% response rate.
The questionnaire was online.

Sexual behaviour in Britain and Australia

Britain conducts a major sexual survey once every 10 years.
The National Survey of Sexual Attitudes and Lifestyles.

Has excellent sampling methods and has a response rate of 58 percent.
They had a very large sample.
Questions were answered on a computer and urine and saliva samples were collected.

A survey of Latino women

Sexual Assault among Latinas (SALAS)

Magazine surveys

Many large-scale sex surveys have been conducted through magazines.
Often the survey is printed in one issue of the magazine and readers are asked to respond.

Sampling is out of control with magazine surveys.
The response rate is also unknown.
Magazine sex surveys are poor in quality because the sample is generally seriously biased.

Studies of special populations

Snowball sampling or respondent-driven sampling: a method for acquiring a sample of people, in which existing participants suggest names of future participants to be recruited.

To interview sexual minorities the interviewers should themselves be sexual minorities.

Web-based surveys

Web-based surveys have many advantages but also some disadvantages.

• Can recruit larger samples than can traditional interview or questionnaire studies
• Can potentially produce broader samples
• It has advantages for studying special populations defined by their sexual behaviour, particularly if the behaviour is taboo
• Have the ability to eliminate extraneous influences on responding
• But, they still rely on self-report
• But, some bias is introduced because not everyone has internet-access
• But, the researcher lacks control of the environment in which the respondent completes the survey

Media content analysis

Content analysis: a set of procedures used to make valid inferences about text.

Many of the same methodological issues discussed earlier come into play with content analysis.

• Sampling
• Create a coding protocol
• The reliability must be demonstrated
  Inter-coder reliability

Qualitative methods

Qualitative research: a collection of naturalistic, holistic methods, including participant observation and in-depth interviewing, in which the results are conveyed not in numbers but in words.

Ethnography: a research method used to provide a description of a human group, a social setting or a society.

Participant-observer technique: a research method in which the scientist becomes part of the community to be studied and makes observations from inside the community.

Experiments

Correlational study: a study in which the researcher does not manipulate variables but rather studies naturally occurring relationships among variables.

Experiment: a type of research study in which one variable is manipulated by the experimenter while all other factors are held constant. The researcher can then study the effects of the independent variable on some measured variable. The researcher can then make causal inferences about the effects of the independent variable on the dependent variable.

Meta-analysis

Meta-analysis: a statistical method that allows the researcher to combine the results of all prior studies on a particular question to see what, taken together, they say.

Incidence: the percentage of people giving a particular response
Frequency: how often a person does something