In psychological evaluation, subjective tests are intended to analyze people’s personality in order to predict possible dysfunctions. In this article we will know one of them, the test of the semantic differential of Osgood and collaborators (1952, 1972).

This test has its theoretical basis in Osgood’s (neo-conductive) media theory, according to which intermediate (covert) cognitive processes modulate the functional relationships between stimuli and responses.

Subjective tests: characteristics

The semantic differential test is classified as a subjective test. The objective of the subjective tests is that the subject describes, classifies or qualifies himself, objects and persons, or that the persons close to the subject do the same with respect to him.

This type of tests are semi-structured , voluntary (that is, the subject can falsify them) and not masked (the subject knows what is being evaluated).

Furthermore, these are non-typified tests; that is, there are no established rules that allow the interpretation of the scores obtained in the test. In relation to this, there are only two exceptions: the ACL (Gough’s Adjective Checklist) and the DACL (Lubin’s Adjective Checklist), which are standardised subjective tests.

Based on the subjective tests, a quantitative or qualitative analysis of the data can be made. Their origin lies in theoretical phenomenological and cognitive approaches, and they are widely used in cognitive-constructivist models.

Semantic differential test: what is it?

The semantic differential test was developed by Charles Osgood, George Suci and Percy Tannenbaum in 1957. This test measures the subjects’ responses to semantic objects or stimuli (known as “concepts”) through scales of estimation defined by opposing bipolar adjectives (e.g., generous/egotistic, distrustful/negligent, nervous/quiet…)

The authors argue that a concept acquires meaning when a sign (word) can provoke the response that is associated with the object it represents; that is, the subject reacts to the object symbolized .

For its construction, the concepts or semantic stimuli are selected through empirical or rational criteria. The test allows to investigate the meaning that the chosen concepts have for a subject or group of subjects.

Formats

The formats of the semantic differential test can be of various types.

For example, one could be the following: it would have as its headline “I ACTUAL”, and underneath the antonyms adjectives in estimation scale format: here the subject must be placed between the adjectives , according to which one or the other is more (increasing the proximity to the adjective that you consider to be the best defining).

Another format would be the one that includes in the header the adjectives antonyms, for example “CARIÑOSO-ARISCO” and underneath the people that the subject will evaluate: “father”, “mother”, “current self” and “partner”, for example.

That is, the subject can evaluate only himself, or evaluate more people (always according to his point of view).

How does it work?

Let’s see in a little more detail how the test unfolds.

A list of adjectives is proposed to the subject, which must be related to the proposed concepts . As we have already seen, the adjectives are presented in bipolar form, with a series of intermediate values between the two extremes. For example, the pair “fair” / “less fair” is presented, separated by a kind of graduated rule in which the subject must mark how he would place the concept in relation to both poles.

It is important to know that concepts of the “good/bad” type should not be contrasted, since the scale of measurement of the semantic differential is not comparative, so questions should always be bipolarized around the same concept.

Factors that saturate the test

The main interest of Osgood and his collaborators was to study the structure of the meaning of the subjects. The authors concluded that such meaning has three dimensions: evaluation, power and activity.

Thus, the estimation scales or bipolar adjectives of the semantic differential test, saturate these three dimensions or factors:

1. Evaluation

This is content that has evaluative connotations (e.g. good/bad; beautiful/ugly).

2. Power

It consists of all content that expresses power or strength (e.g. strong/weak; big/small).

3. Activity

It refers to active content, e.g. slow/fast or passive/active.

Sources of error

There are a number of sources of error in the semantic differential test, coming from the subject(s) being tested. These errors are:

1. Social Desirability

It is the fact of wanting to like or give a good image , on the part of the subject; it influences the evaluative factor.

2. Scalar format

The fact that the semantic differential test is based on estimates from scales means that subjects can present certain response trends, due to the format of the test itself .

Thus, it has been seen how subjects with a high IQ tend to give more central responses on the scale; on the other hand, subjects with a low IQ tend to respond at the extremes. The same happens with depressed subjects (they give central responses) and anxious subjects (they give extreme responses).

Analysis of information

Two types of analysis can be performed in the semantic differential test:

1. Profile analysis

The subject and the opinions he gives about others (for example about his father and mother) are analysed; allows the different scores (of the different subjects) to be compared with each other.

2. Distance analysis

In this case the subject is analyzed at two different time points (“before and after”), although it may include more time points. That is, it allows us to compare the subject’s responses over time, and to observe how the subject has evolved in each of the bipolar adjectives.

Bibliographic references:

  • Cohen, R.J., Swerdlik, M.E. (2002). Psychological testing and assessment. McGraw-Hill. Madrid.
  • Fernández-Ballesteros, R. (2005). Introduction to Psychological Evaluation I and II. Pirámide Publishing House. Madrid.
  • Fernández-Ballesteros, R. (2011) Psychological Evaluation. Concepts, methods and case studies. Pyramid Edition. Madrid.