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

Surveys, Using Others’

Using others’ surveys is a viable option in communication research. If researchers seek to measure a previously studied variable, scholars recommend using or adapting others’ surveys. Typically, existing surveys have been tested, refined, and selected for publication. Using a book such as Communication Research Measures II: A Sourcebook can be a useful starting point when researching existing scales. In this book, the authors provide scales from various communication contexts (e.g., family, health, intercultural). The authors also provide information regarding scale background, reliability, validity, and primary citations. Although taking advantage of others’ surveys is sensible, the simple fact that someone else has developed a survey does not automatically make it a good survey. Thus, making informed decisions about others’ surveys is vital. This entry discusses the process of evaluating others’ surveys, modification of others’ surveys, and finally, advantages of using others’ surveys.

Evaluating Others’ Surveys

Before deciding to utilize a survey, researchers should evaluate whether or not the survey is reliable and valid. The original publication (or follow-up publications evaluating or modifying a survey) in which the survey appears typically has information regarding validity and reliability. Reliability captures the consistency of a survey. Cronbach’s alpha (or the reliability coefficient) is a measure of internal consistency, or how closely related a set of items is as a group. In other words, a reliable survey yields highly similar results each time it is used. Reliability is expressed as a matter of degree using the reliability coefficient, which is a number ranging between 0 (less degree of reliability) and 1 (greater degree of reliability). Typically, researchers report a reliability coefficient when describing their scales. Most communication researchers agree that a reliability coefficient of .70 or above is an acceptable degree of reliability. However, if a construct is more abstract, and thus more difficult to measure, a lower reliability coefficient could be acceptable. Conversely, if a construct is easy to measure, then a reliability coefficient greater than .70 might be expected.

Validity relates to how closely the survey measures what it intends to. When evaluating content or face validity, researchers may ask a question such as, “Are these items representative of the construct being measured?” Researchers can use convergence (i.e., finding significant relationships between measures that should be related, such as relational quality and satisfaction) and/or divergence (i.e., no significant relationship between measures that should not be related, such as introversion and extraversion) procedures to establish validity. Another more advanced technique for evaluating validity is confirmatory factor analysis, which verifies the identity of subscale items as elements of the construct being measured. Before deciding to use a survey, researchers should put in the time and effort to gain information regarding the scale’s reliability and validity. The original source as well as other studies that have used the survey will provide insight into how the survey has performed over time.

Modifying Others’ Surveys

Sometimes a survey needs to be modified by a researcher. For instance, perhaps a commitment scale was developed for married individuals, but the researcher wants to adapt the items for friendships. Rewording the items for friendships, as opposed to marriages, is not too daunting. For instance, an item that states, “I am very committed to my partner” can be easily adjusted to, “I am very committed to my friend.” However, the researcher needs to be cautious because this adjustment could change the intent of the scale, along with its validity and reliability. Pretesting the adjusted scale (which will be discussed in greater detail later in this entry) will help the researcher determine if the scale captures what is intended. Similarly, translating a survey to or from English potentially poses issues, as certain words are more difficult to translate (e.g., different meanings in other languages), some cultures are less willing to share personal information, and so on. When survey translation is necessary, researchers can collaborate with native speakers of a language to make sure the intent of the questions is clear. After translating a survey, pretesting techniques involving feedback from bilingual respondents are also useful.

...

  • 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