Information from Keywords

Alvarez P., Pulgarin A. (1996) Information from keywords. Rasch Measurement Transactions 10:2 p. 493.

The bibliographic value of keywords is problematic. In an attempt to learn from them, Alvarez and Pulgarin (1996) apply Rasch analysis to a subset of keywords appearing in the journal Diabetes from 1974 to 1994. The plot is derived from numbers published in Table 2 of their paper. High measures indicate that the keyword appeared with high frequency across the years. The most frequent keyword is "Diabetes m. exp." (Diabetes mellitus experimental) which was highly frequent from 1981 to 1993. The keyword that is least predictable across the years according to the reported fit statistic is "Diabetes m. non-ins. dep." (Diabetes mellitus non-insulin dependent. But this keyword shows a predictable pattern in the data. It was not used prior to 1983, but has been increasing steadily in use since then. Many keywords are used rarely. Metformin is so predictable because it is only used once or not at all in each year.

These results show that, even with this awkward data, imposing a measurement framework enables one to learn. In fact, it encourages one to reconceptualize the problem such that each word has a measure for each year. Then the trend across time for each word can be plotted to reveal both substantive changes in research emphasis and ephemeral fashions in keyword use.

Alvarez P., Pulgarin A. (1996) The Rasch Model. Measuring information from Keywords: The Diabetes Field. Journal of the American Society for Information Science 47:6 468-476.

Alvarez P., Pulgarin A. (1996) Information from keywords. Rasch Measurement Transactions 10:2 p. 493.

Information from keywords. Alvarez P., Pulgarin A. … Rasch Measurement Transactions, 1996, 10:2 p. 493

The URL of this page is www.rasch.org/rmt/rmt102a.htm

Website: www.rasch.org/rmt/contents.htm