Compound Linguistic Scale

•Compound Linguistic Scale comprises Compound Linguistic Variable, Fuzzy Normal Distribution and Deductive Rating Strategy.•CLV can produce two dimensional options, i.e. compound linguistic terms, to better reflect the raters’ preferences.•DRS is a double step rating approach for a rater to choose a compound linguistic term among two dimensional options.•FND can efficiently produce a population of fuzzy numbers for a linguistic term set with using a few parameters.•CLS, as a rating interface, can be contributed to various application domains in engineer and social sciences.

Rating scales are the essential interfaces for many research areas such as in decision making and recommendation. Some issues concerning syntactic and sematic structures are still open to discuss. This research proposes a Compound Linguistic Scale (CLS), a two dimension rating scale, as a promising rating interface. The CLS is comprised of Compound Linguistic Variable (CLV) and Deductive Rating Strategy (DRS). CLV can ideally produce 21 to 73 ((7 ± 2)((7 ± 2) − 1) + 1) ordinal-in-ordinal rating items, which extends the classic rating scales usually on the basis of 7 ± 2 principle, to better reflect the raters’ preferences whilst DRS is a double step rating approach for a rater to choose a compound linguistic term among two dimensional options on a dynamic rating interface. The numerical analyses show that the proposed CLS can address rating dilemma for a single rater and more accurately reflects consistency among various raters. CLS can be applied to surveys, questionnaires, psychometrics, recommender systems and decision analysis of various application domains.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slide