The Guttman scale (named after Louis Guttman, eng.Guttman scale / scalp gram) is a measuring scale based on one-dimensionality and the fact that the statements in the questionnaire are hierarchically related. Scalogram analysis data based on the respondents' answers (in the "yes - no" format) are compared with the test or questionnaire data. All who answered positively in relation to the main question will be located at the beginning of the scale, and negatively - at the end. If one of the survey judgments receives more than 80% of positive or negative responses, it is thrown out. [one]
Content
Appearance history
Guttman pursued a goal in constructing uniform scales with the help of cumulative (accumulative) statements. He criticized the existing scales for their “multidimensionality” and believed that if we have a number of indicators, then a certain sequence is observed in them. Guttman's analysis of the grams suggested that the points were arranged on a scale in such a way that, depending on the answers, all the respondents could be placed on it in one row. This means that the respondent’s setup score is uniquely displayed on the scale. All respondents positively to this survey are placed at the beginning of the series, and negatively - at the end.
Scale Highlights
Method Guttman consists of eight stages. First stage. Selection of judgments. The researcher a priori chooses judgments relating to the installation of people on the object or phenomenon under study. Whether these judgments can form a scale is determined by expert verification. Requirements for judgments: - the presence of two alternative answers - only "yes" or "no" - focus on the study of any one specific area - location on a cumulative (accumulative) scale, i.e. if the respondent gave a positive response to the first judgment, then there is a high probability that he will give positive answers to the subsequent ones.
Second phase. Selection of experts.
The third stage. Expert assessment of the proposed judgments.
Fourth stage. Data processing.
Fifth stage. Checking the scale for reproducibility. Judgments are considered suitable for scaling if the reproducibility factor is 0.85. It characterizes the degree of proximity to the ideal scale in which the reproducibility is 1.00.
The sixth stage. Selection of reliable judgments. If the reproducibility is less than 0.85, then judgments that give a lot of erroneous answers are discarded. A new scale log is being built and a new reproducibility factor is being calculated. This procedure is repeated until reliable and meaningful judgments are selected. A scale with a reproducibility factor of at least 0.85 is ready.
Seventh stage. Survey of respondents. In a mass survey, all judgments are shuffled in disarray.
The eighth stage. Mathematical processing of the data. The rank of each respondent is determined by the sum of his points. Then the arithmetic average rank of this category of the examined is calculated. Compared with the ranks of other categories. For example, the ranks of representatives of various professions, demographic, ethnic and other groups, to carry out correlation, factor analysis and so on. The average category rank shows the overall installation of this group of people on the object or phenomenon being studied. [2] The advantages of the scale are: 1) the one-dimensionality of the scale is based on rigorous methods 2) the measurement is quite simple
Usage Example
You can consider the Guttman scale on the example of constructing a scale to measure people's social attitudes regarding the transition to a new system of work organization. The questionnaire contains several judgments in which respondents choose the answer “Agree” or “Disagree”. For a particular opinion is given 1 point. In the example below, agreement in options 1, 2, 5, 6 and disagreement in options 3, 4, 7, 8 means the most favorable attitude in the new system of work organization.
1. The new system of organization undoubtedly contributes to increased productivity. Agree (1) - Disagree (0)
2. In general, this system is better than the one used before. Agree (1) - Disagree (0)
3. Some aspects of the new organization system are poorly thought out. Agree (0) - Disagree (1)
4. Like any other system of organization, the new system has many disadvantages. Agree (0) - Disagree (1)
5. The new system successfully combines material and moral incentives for workers. Agree (1) - Disagree (0)
6. The case for the new system is very convincing. Agree (1) - Disagree (0)
7. In the former system there was a lot of good that was lost in the new organization. Agree (0) - Disagree (1)
8. The advantages of the new organization system are completely unclear. Agree (0) - Disagree (1)
In an ideal scale, the answer to one of the points entails a definite answer to the next. That is, it is important to make a questionnaire so that the answers form a one-dimensional continuum. If everyone positive for a polling answer gives 1 point, and negative - 0, then supporters of the idea (in this case, a new system of work organization) will receive 8 points, and opponents of the idea - 0. And the rest will be located on a scale between them, depending on from their answers. [3]
Criticism
The Guttman scale addresses a number of serious criticisms, the main one being the impossibility of constructing a one-dimensional scale in many cases (for example, when measuring artistic preferences). The disadvantages of the scale also include large technical and theoretical difficulties associated with its construction. In addition, the scale may be one-dimensional for one group of individuals and not be such for another. [2]
Literature
Tatarova GG, Methodology of data analysis in sociology (introduction) / Textbook for universities. - M .: NOTA BENE, 1999. - 224 p.
Yadov VA, Sociological research: methodology, program, methods - Moscow: Nauka, 1972
Gutman L. The main components of the school analysis // Mathematical methods in modern bourgeois sociology. M „1966
Torgerson WS Theory and methods of scaling. NY, 1958
Koppen MGM On finding the bidimension of a relation // J. of math Psych. 1987. VoL 31, No. 2. G.A. Satarov
Burganova, IN, The Theory of Measurement in Sociology: A Teaching Aid for Bachelors
Notes
- Burganova IN, The Theory of Measurement in Sociology: A Teaching Aid for Bachelors
- ↑ 1 2 Tatarova GG, Methodology of data analysis in sociology (introduction) / Textbook for universities. - M .: NOTA BENE, 1999. - 224 p.
- ↑ Yadov VA, Sociological research: methodology, program, methods - Moscow: Nauka, 1972.