Self Report In Research and Use of Likert Scale
The most widely used scaling technique is the Likert scale, named after the psychologist Rensis Likert. A Likert scale consists of several declarative items that express a viewpoint on a topic. Respondents are asked to indicate the degree to which they agree or disagree with the opinion expressed by the statement. Some researchers prefer seven-point scales, adding the alternatives “slightly agree” and “slightly disagree.”
There are also diverse opinions about including an explicit “uncertain” category. Some argue that this option makes the task more acceptable to people who cannot make up their minds or have no strong feelings about an issue. Others, however, feel that an undecided category encourages fence-sitting, that is, a tendency not to take sides.
After subjects complete a Likert scale, their responses are scored. Typically, agreement with positively worded statements and disagreement with negatively worded statements are assigned higher scores. Thus, a higher score would be assigned to those agreeing with this statement than to those disagreeing with it.
With five response alternatives, a score of 5 would be
given to those strongly agreeing, 4 to those agreeing, and so forth. The
responses of two hypothetical respondents are shown by a check or an X, and
their scores for each item are shown in far right columns.
The summation feature of such scales makes it possible to make fine discriminations among people with different points of view. Although traditional Likert scales were used to measure attitudes, summated rating scales can be used to measure a wide array of attributes.
In such cases, the bipolar scale would not be
agree/disagree but might be always true/never true, extremely likely/extremely
unlikely, and so on.
Semantic Differential Scales
Another technique for measuring psychosocial traits is the semantic differential (SD). With the SD, respondents are asked to rate a concept (eg, primary nursing, team nursing) on a series of bipolar adjectives, such as effective/ineffective, good/bad, important/unimportant, or strong/weak.
Respondents place a check at the appropriate point on seven-point rating scales extending from one extreme of the dimension to the other. Semantic differentials are flexible and easy to construct. The concept being rated can be virtually anything a person, place, situation, abstract idea, controversial issue, and so forth. The concept can be presented as a word, as a phrase, or even as visual material ( eg photos, drawings).
Typically, several concepts are included on an SD so that comparisons can be made across concepts (eg, male nurse, female nurse, male physician, female physician). Researchers also have a leeway in selecting the scales, but two considerations should guide the selection. First, the adjective pairs should be appropriate for the concepts being used and for the information being sought.
The second consideration in selecting adjective pairs is the extent to which the adjectives measure the same dimension of the concept. Osgood, Suci , and Tannenbaum (1957), through extensive research with SD scales, found that adjective pairs tend to cluster along three independent dimensions: evaluation, potency, and activity.
The most important group of adjectives are evaluative ones, such as effective/ineffective, valuable/worthless, good/bad, fair/unfair, and so forth. Potency adjectives include strong/weak and large/small, and examples of activity adjectives are active/passive and fast/slow. These three dimensions need to be considered separately because subjects' evaluative ratings of a concept are independent of their activity or potency ratings.
For example, two people who associate high levels of activity with the concept nurse practitioner might have divergent views about how to evaluate nurse practitioners. Researchers must decide whether to represent all three dimensions or whether only one or two are needed.
Each dimension must be
scored separately. Scoring of SD responses is essentially the same as for
Likert scales. Scores from 1 to 7 are assigned to each bipolar scale response,
with higher scores usually associated with the positively worded adjective.
Responses are then summed across the bipolar scales to yield a total score.
Existing Self-Report Scales and Psychological Measures
Clinical nurse researchers have
studied many psychosocial traits, and numerous self-report scales have been
developed to measure them, often using a summated rating scale format. Note that many scales and test
instruments must be purchased from the publisher, or require the author's
permission to use them.
The CINAHL database includes information on the scales
used in research studies. Information on standardized tests and psychological
measures can be retrieved through a computerized literature search of the
database called Mental Measurement Yearbook, produced by the Buros Institute of
Mental Measurements, or through Health and Psychosocial Instruments
Give your opinion if have any.