Language Insights to Predict Behavior

By seedlink In Research Posted Wed 06 April 2016

Can language use reflect personality style? This is the question two professors, James W. Pennebaker of University of Texas at Austin and Laura A. King of Southern Methodist University, had been asking. They conducted multiple studies to examine the reliability, factor structure, and validity of written language using a word-based, computerized text analysis program. In the subsequent academic paper Linguistic Styles: Language Use as an Individual Difference they concluded that linguistic style is an independent and meaningful way of exploring personality. This article is a simplified outline of the academic paper, extracting and distilling the major points from the original content.

Linguistic analysis at a glimpse

Scientific research conducted by the two professors proves that linguistic style can be used as an accurate way to analyze personality, as everyone has a distinctive way of expressing themselves with language. Since no human sees the world the same way the others do, how they talk about things can subconsciously reveal a pattern and offer us an insight into their unique perception.

In order to harness the organic way to assess people and their personality, researchers developed a text analytic program called LINGUISTIC INQUIRY AND WORD COUNT (LIWC). What is so powerful about this program is that it provides a systematic and objective method of assessing individuals in the most efficient manner, and more importantly, does not compromise on accuracy of its results at all.

pic

The origins of using language to analyze human traits

The role of personality in the prediction of behavior is evident in the stable consistency of language use across a variety of written contexts. In other words, regardless of the contents that an individual has written about different topics, the way that individual expresses himself (via phrasing, thought units and word choices etc.) remains the same. As the way people talk about things subconsciously reveals important information about their personality, verbal expression is an essential source of data about individuals.

The idea of studying multiple psychological dimensions of speech and text was conceived in the medical field to understand the underlying emotional cognitive dynamics by comparing the use of linguistic dimensions across people with different medical diagnoses and historical personalities (Gottschalk & Gleser, 1969), or to the form and intent of disclosures via verbal response mode by transcribing more natural interactions between physician and patient (Stiles, 1992). It was heavily psychoanalytic in orientation and required trained rates to evaluate each clause of a sentence via content analysis.

Initial problems of linguistic analysis program

There were several problems that plagued the initial adoption of word usage approach to the analysis of naturally occurring written or spoken language. First, content analysis requires the additional resources of human raters to make sense of writing or speech sample to recognize the underlying motives. Moreover, when reading for meaning raters may fail to recognize the patterns of word usage over different topics of writing.

The other alternative to content analysis would be word-base count text analysis, which is based on the hypothesis that one or more general concepts that a person is attempting to express can be captured by examining the specific words used to make up the concepts. For example, someone who is expressing sadness would be more likely to use words such as “sad”, “cry”, “loss” or “alone”. The drawback of this word-based counting system toward analyzing human qualities is the crippling inability to consider context, irony, sarcasm or multiple meanings of words. As such, the accuracy of the results would be relatively questionable if the true intentions of the writer failed to be picked up by the program.

pic
Professor James W. Pennebaker of University of Texas at Austin, together with Laura A. King of Southern Methodist University, is the co-author of the research paper Linguistic Styles: Language Use as an Individual Difference.

Innovative technological breakthrough: LIWC Program

There was a great concern over the development of efficient ways to examine a large number of text files to learn whether the words used to express oneself can predict long term psychological and physical health. With the above issues in mind, renowned researchers all worked together to seek a system that would provide a wide range of psychological dimensions, and therefore came up with a word-based text analysis program called LINGUISTIC INQUIRY AND WORD COUNT (LIWC). Unlike conventional text analysis strategies where the assignment of value of the words is made within a specific context, LIWC’s subjective dictionaries were independently rated by judges. The language categories used in LIWC were carefully constructed, subsequently validated by having judges rate hundreds of files written text and undergone vigorous cross-reference to establish credibility of the program. Using LIWC, investigation was conducted to determine the degree to which language use is reliable across time and topic, possesses a reliable factor structure, and exhibits good construct and divergent validity in comparison with established personality measures.

LIWC relied on thematic analysis of imaginative stories as a means to reveal the non-conscious motives that orient attention and drive behavior. The idea that the way people express themselves in language provides psychologically important information is a central assumption of the LIWC strategy presented, and is also proven to be quite reliable across topics and testing occasions. LIWC variables were correlated with the five-factor measure of personality, various demographic measures, several health and health-related behavior markers and assorted commonly used self-report personality instruments. In other words, by tapping into the very way in which people are cognitively organizing their language as reflected through the word choice in their reports, we are able to gain an accurate insight into their personality traits and even demographic variables through the language used.

Drawing to the conclusion

The findings support the idea that reliable linguistic styles can be identified and even can help provide an understanding of how individuals convey their thoughts and feelings. The insight into the person’s general approach to the word through analysis of the emergence of characteristic speech patterns.

Linguistic strategy offers another comprehensive perspective in tapping individual difference styles – the study of personality can be conducted by isolating reliable language patterns from person to person. This innovative way of personality assessment is not only comparatively more objective in the sense that it eliminates human error, it is also an organic way to open a window to understand the ways people perceive the world and naturally express themselves.

How Seedlink uses this technology to contribute to corporate success

Here at Seedlink, we believe that the most valuable asset in a company is its people, and it is paramount to find the right people, who have not only the necessary technical skills, but also the ability to interact effectively and harmoniously with other people. Seedlink utilizes this ground-breaking technology by applying it to the recruitment industry. Job seekers are asked to answer some open-ended questions, and from the data we can confidently crunch the data to sieve out only the best candidates who possess crucial soft skills and most fit the company culture.

See how Seedlink's A.I. can help you scale your business.

Send me the latest news in HR technology
Keep me posted on sharing sessions and events
Get a free demo of Seedlink's A.I.

RECOMMENDED ARTICLES

By seedlink In Research Posted 09 一月 2016
2017 HR趋势 | 提供4E式的领导力培养
By seedlink In Research Posted 09 一月 2016
2017 HR趋势 | 产品思维将成为HR部门的主流
By seedlink In Research Posted 09 一月 2016
2017 HR趋势 | 实时反馈工具和People Analytics
By seedlink In Research Posted 09 一月 2016
2017 HR趋势 | 不仅是扁平化,更要小团队化
By seedlink In Research Posted Wed 06 July 2016
人工智能的白种人难题
By seedlink In Research Posted Mon 27 June 2016
Same personality? It can’t be true!
By seedlink In Research Posted Mon 20 June 2016
With 3 simple questions, you can leverage A.I. to recruit cream of the crop
By seedlink In Business Posted Fri 03 June 2016
Winning Award As HR Service Vendor with Best Potentials
By seedlink In Business Posted Wed 04 May 2016
Seedlink Winning Awards on International Women Venture Day