Semantic similarity scales : Using semantic similarity scales to measure depression and worry
Kjell, Oscar N.E. LU ; Kjell, Katarina LU ; Garcia, Danilo and Sikström, Sverker LU
(2020)
p.53-72
- Abstract
The aims of this chapter include describing: how the semantic representations may be used to measure the semantic similarity between words. the validity of semantic similarity as measured by cosine. how semantic similarity scales can be used in research. how to apply t-test to compare two sets of texts using semantic similarity (i.e. “semantic” t-test). how to visualize the word responses by plotting words according to semantic similarity scales. a research study where depression is measured using semantic similarity scales, independent from traditional rating scales. This chapter describes how semantic representations based on Latent Semantic Analysis (LSA; Landauer and Dumais 1997) may be used to measure the semantic similarity... (More)
The aims of this chapter include describing: how the semantic representations may be used to measure the semantic similarity between words. the validity of semantic similarity as measured by cosine. how semantic similarity scales can be used in research. how to apply t-test to compare two sets of texts using semantic similarity (i.e. “semantic” t-test). how to visualize the word responses by plotting words according to semantic similarity scales. a research study where depression is measured using semantic similarity scales, independent from traditional rating scales. This chapter describes how semantic representations based on Latent Semantic Analysis (LSA; Landauer and Dumais 1997) may be used to measure the semantic similarity between two words, sets of words or texts. Whereas Nielsen and Hansen describe how to create semantic representations in Chap. 1; this chapter focuses on describing how these may be used in research to estimate how similar words/texts are in meaning as well as testing whether two sets of words statistically differ. This approach may, for example, be used to detect between group differences in an experimental design. First, we describe how a single word’s semantic representation may be added together to describe the meaning of several words or an entire text. Second, we discuss how to measure semantic similarity using cosine of the angle of the words’ position in the semantic space. Third, we describe how this procedure of text quantification makes it possible for researchers to use statistical tests (e.g., semantic t-test) for investigating, for example, differences between freely generated narratives. Lastly, we carry out a research study building on studies by Kjell et al. (2018) that demonstrated that semantic similarity scales may be used to measure, differentiate and describe psychological constructs, including depression and worry, independent from traditional numerical rating scales.
(Less)
- author
- Kjell, Oscar N.E.
LU
; Kjell, Katarina
LU
; Garcia, Danilo
and Sikström, Sverker
LU
- organization
- publishing date
- 2020
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- Statistical Semantics : Methods and Applications - Methods and Applications
- pages
- 20 pages
- publisher
- Springer International Publishing
- external identifiers
-
- scopus:85089334278
- ISBN
- 9783030372507
- 9783030372491
- DOI
- 10.1007/978-3-030-37250-7_4
- language
- English
- LU publication?
- yes
- id
- a578e1d9-600b-462b-a545-c2ad7fa58403
- date added to LUP
- 2020-08-24 08:51:17
- date last changed
- 2024-06-12 19:17:24
@inbook{a578e1d9-600b-462b-a545-c2ad7fa58403,
abstract = {{<p>The aims of this chapter include describing: how the semantic representations may be used to measure the semantic similarity between words. the validity of semantic similarity as measured by cosine. how semantic similarity scales can be used in research. how to apply t-test to compare two sets of texts using semantic similarity (i.e. “semantic” t-test). how to visualize the word responses by plotting words according to semantic similarity scales. a research study where depression is measured using semantic similarity scales, independent from traditional rating scales. This chapter describes how semantic representations based on Latent Semantic Analysis (LSA; Landauer and Dumais 1997) may be used to measure the semantic similarity between two words, sets of words or texts. Whereas Nielsen and Hansen describe how to create semantic representations in Chap. 1; this chapter focuses on describing how these may be used in research to estimate how similar words/texts are in meaning as well as testing whether two sets of words statistically differ. This approach may, for example, be used to detect between group differences in an experimental design. First, we describe how a single word’s semantic representation may be added together to describe the meaning of several words or an entire text. Second, we discuss how to measure semantic similarity using cosine of the angle of the words’ position in the semantic space. Third, we describe how this procedure of text quantification makes it possible for researchers to use statistical tests (e.g., semantic t-test) for investigating, for example, differences between freely generated narratives. Lastly, we carry out a research study building on studies by Kjell et al. (2018) that demonstrated that semantic similarity scales may be used to measure, differentiate and describe psychological constructs, including depression and worry, independent from traditional numerical rating scales.</p>}},
author = {{Kjell, Oscar N.E. and Kjell, Katarina and Garcia, Danilo and Sikström, Sverker}},
booktitle = {{Statistical Semantics : Methods and Applications}},
isbn = {{9783030372507}},
language = {{eng}},
pages = {{53--72}},
publisher = {{Springer International Publishing}},
title = {{Semantic similarity scales : Using semantic similarity scales to measure depression and worry}},
url = {{http://dx.doi.org/10.1007/978-3-030-37250-7_4}},
doi = {{10.1007/978-3-030-37250-7_4}},
year = {{2020}},
}