Skip to main content icon/video/no-internet
icon backReturn to Entries

Representativeness

This entry discusses the concept of representativeness and its importance in social and behavioral science research. In conducting such research, it is desirable but usually not possible to study an entire population; in consequence, researchers sample from the population and then generalize findings of the research based on the sample to the population. The analytic sample is thus supposed to represent the population. It is important to know whom a study sample represents to evaluate implications of the study and findings. Who is represented in a study’s sample frames the study.

Two Main Types of Representativeness

The best samples are called probability samples; these incorporate some form of random selection into their sampling procedure, so it is straightforward to discern whom the sample represents. By design, a probability sample is a random sample of some specified target population, and researchers (as well as consumers of the research) can be confident that findings generated by a probability sample represent or generalize to that specified target population. By contrast, convenience samples entail collecting data from samples in an ad hoc or “first-come, first-served” basis. Such samples usually do not represent some larger population, and so it is not possible to apply a convenience sample’s findings to a broader population beyond the specific individuals sampled.

Depending on their target populations, probability samples vary in the scope or degree of representativeness. Consider two probability samples: one of adults aged 55 or older who reside in New York City and another aged 55 or older who reside in the United States. Because both studies are based on probability samples, determining whom each study’s findings represent is clear and straightforward. However, the probability sample of New York City represents a narrower and more restricted population, whereas the probability sample of the United States clearly represents a broader and more diverse population.

Threats to Representativeness

There are different kinds of threats to representativeness. Convenience sampling is one. Attrition is another. Attrition occurs when respondents drop out of a study and/or are lost over time. Attrition is common with longitudinal designs and is a threat to a study’s representativeness because attrition is typically nonrandom (certain types of study participants, such as those from lower socioeconomic backgrounds, are more likely to drop out of a study than others).

To adjust for the effects of attrition, researchers can take advantage of “missing data” approaches.

Sample Weights

To ensure that a sample represents a diverse population and that sufficient numbers are included in a sample for the purpose of statistical power, researchers sometimes purposefully oversample a subpopulation. Then “sample weights” are applied to the data that result in parameter estimates (i.e., estimates of totals, proportions, and associations) that render the data representative of the population from which the sample was recruited.

See also Convenience Sampling; Longitudinal Data Analysis; Validity; Weighting

Marc H. Bornstein Justin Jager
http://dx.doi.org/10.4135/9781506326139.n588
10.4135/9781506326139.n588

Further Readings

Bornstein, M. H., Jager J., & Putnick, D. L. (2013). Sampling in developmental science: Shortcomings and solutions. Developmental Review, 33(4), 357–370.
Davis-Kean, P. E., & Jager,

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Read next

More like this

Sage Recommends

We found other relevant content for you on other Sage platforms.

close

Have you created a personal profile? Login or create a profile so that you can save clips, playlists and searches

close
close
close Signed In to profile You are signed in as:
close
Logged in Institution Institution