Measuring emotional expression with the Linguistic Inquiry and Word Count

The Linguistic Inquiry and Word Count (LIWC) text analysis program often is used as a measure of emotion expression, yet the construct validity of its use for this purpose has not been examined. Three experimental studies assessed whether the LIWC counts of emotion processes words are sensitive to verbal expression of sadness and amusement. Experiment 1 determined that sad and amusing written autobiographical memories differed in LIWC emotion counts in expected ways. Experiment 2 revealed that reactions to emotionally provocative film clips designed to manipulate the momentary experience of sadness and amusement differed in LIWC counts. Experiment 3 replicated the findings of Experiment 2 and found generally weak relations between LIWC emotion counts and individual differences in emotional reactivity, dispositional expressivity, and personality. The LIWC therefore appears to be a valid method for measuring verbal expression of emotion.     

"A picture's worth a thousand words": language use in the autophotographic essay

The authors applied Pennebaker's Linguistic Inquiry and Word Count (LIWC) program to autophotographic essays addressing the question, "Who are you?" to test hypotheses about individuality and social connectedness. Specifically, the authors tested whether insight-oriented words would correlate with individuality ratings, defining essays that portray unique persons who creatively explore the self. Second, the authors tested whether connectedness, operationalized by photos of people-touching and self with others, would correlate with social process words. In 164 student photo essays, greater linguistic cognitive complexity characterized individuality, even when controlling for needs for cognition and uniqueness, and verbal intelligence. Connectedness predicted social words. These findings show that autophotography essays do capture complex self-reflective thought and suggest LIWC may be a valuable tool for understanding the expression of individuality and relatedness.     

Text messaging, personality, and the social context

The purpose of this research was to undertake some analyses of how the language used in text messaging varies as a function of personality traits and the interpersonal context. After completing personality questionnaires, participants provided their most recent text messages and indicated their relationship with the message recipient on several dimensions. Correlations between Linguistic Inquiry and Word Count (LIWC) categories and personality traits and relationship status were examined. There were significant correlations between certain LIWC categories and extraversion (e.g., personal pronouns), neuroticism (e.g., negative emotion words) and agreeableness (e.g., positive emotion words), suggesting that personality traits are displayed in how one texts. One of the defining features of texting - linguistic alterations (e.g., abbreviations) - varied as a function of both personality traits and relationship status. Overall, the results provide a snapshot of what text messages look like, and how they reflect the texter’s personality and the interpersonal context.     

Positive emotions, emotional intelligence, and successful experiential learning

This study explored the role that positive emotions and emotional intelligence play in experiential learning. Students’ field practicum journals were analyzed using the Linguistic Inquiry and Word Count Program (LIWC) and a measure of emotional intelligence was obtained using the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT). Positive emotion words were robustly associated with almost all dimensions of supervisors’ ratings of students’ performance, but showed weak associations with students’ ratings of perceived benefits associated with their practicum experiences. Overall EI scores were correlated with several of the supervisor rating items and the Facilitating Thought and Managing Emotions subscales of the EI were robustly correlated with many of the student rating items. This study thus yielded a more differentiated view of the role of positive emotions and emotional intelligence in adaptive functioning and underscored the importance of using multiple informants to assess a complex construct such as successful experiential learning.     

The Bitter-Sweet Labor of Emoting: The Linguistic Comparison of Writers and Physicists

ABSTRACT: In this exploratory archival study, the motivation of writers of fiction and physicists was examined by studying word usage as a clue to unconscious motivators of their work. The hypothesis was that artists make art to deal with issues in their own lives, thus relying on emotions, particularly negative emotions (markers of presence of issues), to govern their work. Consequently, it was predicted that distinguished writers of fiction, as compared to distinguished physicists, would use more emotion-related words when discussing their work, particularly negative emotion-related words. Interviews conducted with 9 physicists were matched to the interviews with 9 writers, and analyzed using the Linguistic Inquiry and Word Count (LIWC) program (Pennebaker, Francis, & Booth, 2001). Writers used significantly more emotion-related words, in particular more negative-emotion words, including the greater use of anger-related, anxiety-related, and depression or sadness-related words. Almost identical results were obtained when the 9 physicists were compared to the nonmatched, larger sample of 124 writers. The study implies differences of inner preoccupation (relating to work) between creative people oriented towards literary art and physical science.     

Linguistic correlates of asymmetric motor symptom severity in Parkinson’s Disease

Asymmetric motor severity is common in Parkinson’s Disease (PD) and provides a method for examining the neurobiologic mechanisms underlying cognitive and linguistic deficits associated with the disorder. In the present research, PD participants (N = 31) were assessed in terms of the asymmetry of their motor symptoms. Interviews with the participants were analyzed with the Linguistic Inquiry and Word Count (LIWC) program. Three measures of linguistic complexity – the proportion of verbs, proportion of function words, and sentence length – were found to be affected by symptom asymmetry. Greater left-side motor severity (and hence greater right-hemisphere dysfunction) was associated with the production of significantly fewer verbs, function words, and shorter sentences. Hence, the production of linguistic complexity in a natural language context was associated with relatively greater right hemisphere involvement. The potential neurobiological mechanisms underlying this effect are discussed.     

The “Deliberative Digital Divide:” Opinion Leadership and Integrative Complexity in the U.S. Political Blogosphere

This study examined the association between political ideology and linguistic indicators of integrative complexity and opinion leadership in U.S. political blog posts (N = 519). Using Linguistic Inquiry and Word Count (LIWC) text analysis, we found that the posts of conservative bloggers were more integratively simple than those of liberal bloggers. Furthermore, in support of a proposed opinion leadership model of integrative complexity, the relationship between ideology and integrative complexity was mediated by psychological distancing (an indicator of a hierarchical communication style). These findings demonstrate an ideological divide in the extent to which the blogosphere reflects deliberative democratic ideals.     

The language used in describing autobiographical memories prompted by life period visually presented verbal cues,event-specific visually presented verbal cues and short musical clips of popular music

Here, we examined linguistic differences in the reports of memories produced by three cueing methods. Two groups of young adults were cued visually either by words representing events or popular cultural phenomena that took place when they were 5, 10, or 16 years of age, or by words referencing a general lifetime period word cue directing them to that period in their life. A third group heard 30-second long musical clips of songs popular during the same three time periods. In each condition, participants typed a specific event memory evoked by the cue and these typed memories were subjected to analysis by the Linguistic Inquiry and Word Count (LIWC) program. Differences in the reports produced indicated that listening to music evoked memories embodied in motor-perceptual systems more so than memories evoked by our word-cueing conditions. Additionally, relative to music cues, lifetime period word cues produced memories with reliably more uses of personal pronouns, past tense terms, and negative emotions. The findings provide evidence for the embodiment of autobiographical memories, and how those differ when the cues emphasise different aspects of the encoded events.     

Preliminary analysis of language styles in a sample of schizophrenics

Language use of schizophrenics and normal people was compared by applying the language analysis program, Korean Linguistic Inquiry and Word Count. Participants were asked to write a story about the most emotional experience of their lives on A4 size paper. 28 schizophrenics (M age: 26 yr.) and 32 normal people (M age: 23 yr.) participated. Analysis showed normal people used more words about jobs and achievements and fewer words about sex and food. The schizophrenics used fewer pronouns, personal pronouns, and adverbs than the normal group. Some aspects of clinical mechanism are manifest in language uses.     

Does exposure to ambient odors influence the emotional content of memories?

This study examined the effects of implicitly presented odors on the emotional content of memory narratives. Participants were exposed to low levels of perfumes or no odors while they reconstructed a memory from childhood and a recurrent dream. Narratives were analyzed for emotional content using the Linguistic and Inquiry Word Count program. Exposure to perfumes increased the percentage of positive emotion words used by participants in recalling both dreams and childhood memories. In addition, exposure to odors decreased the percentage of negative emotion words that men used in their narratives, whereas the opposite effect was found for women. Discussion focuses on gender differences in written emotional expression, possible ways by which perfumes may exert their effects, and the usefulness of using ambient odors and objective content analysis when studying emotions in memories.     

Caregivers’ Positive Emotional Expression and Children’s Psychological Functioning after Parental Loss

The death of a loved one, particularly a parent, has been identified as not only the most common, but also the most distressing form of adversity youth may experience in their lifetime. Surviving caregivers’ communication with their children may play a critical role in shaping bereaved children’s psychological functioning. However, few studies have examined the specific content (e.g., word usage) of caregivers’ verbal communication as a predictor of psychological functioning in bereaved youth. In a sample of 39 parentally-bereaved children and their surviving caregivers, we investigated whether the frequency of caregivers’ use of positive emotion words (e.g., “love”, “happy”, “hope”) during a reminiscing task about the deceased was associated with children’s psychological functioning and coping. In a cross-sectional analysis, we specifically examined whether these associations were moderated by the amount of time passed since children lost their parents. The Linguistic Inquiry and Word Count Program (LIWC) was used to code and evaluate the percentage of positive emotion words caregivers used during the discussion. When caregivers used more positive emotion words, children were less likely to experience depression, anxiety, and avoidant coping. Those associations were present for children who had experienced parental loss at least 105 days prior to the study.     

A text analysis of the poems of Sylvia Plath

Changes in the words used in the poems of Sylvia Plath were examined using the Linguistic Inquiry and Word Count, a computer program for analyzing the content of texts. Major changes in the content of her poems were observed over the course of Plath's career, as well as in the final year of her life. As the time of her suicide came closer, words expressing positive emotions became more frequent, while words concerned with causation and insight became less frequent.     

The big two dictionaries: Capturing agency and communion in natural language

Four studies developed and validated two dictionaries to capture agentic and communal expressions in natural language. Their development followed the Linguistic Inquiry and Word Count (LIWC) approach (Study 1) and we tested their validity with frequency-based analyses and semantic similarity measures. The newly developed Agency and Communion dictionaries were aligned with LIWC categories related to agency and communion (Study 2), and corresponded with subjective ratings (Study 3), confirming their convergent validity. Very low or absent correspondence between proposed dictionaries and unrelated LIWC categories demonstrated their discriminant validity (Study 2). Finally, we applied both dictionaries to language used in advertisements. In correspondence to gender stereotypes, male-dominated jobs were advertised with more agentic than communal words, and female-dominated jobs with more communal than agentic words (Study 4). Both dictionaries represent reliable tools for quantifying agentic and communal content in natural language, and will improve and facilitate future research on agency and communion.     

A Comparison of Human Narrative Coding of Redemption and Automated Linguistic Analysis for Understanding Life Stories

The majority of life narrative research is performed using trained human coders. In contrast, automated linguistic analysis is oft employed in the study of verbal behaviors. These two methodological approaches are directly compared to determine the utility of automated linguistic analysis for the study of life narratives. In a study of in‐person interviews (N = 158) and a second study of life stories collected online (N = 242), redemption scores are compared to the output of the Linguistic Inquiry and Word Count (Pennebaker, Francis & Booth, 2001). Additionally, patterns of language are found using exploratory principal components analysis. In both studies, redemption scores are modestly correlated with some LIWC categories and unassociated with the components. Patterns of language do not replicate across samples, indicating that the structure of language does not extend to a broader population. Redemption scores and linguistic components are independent predictors of life satisfaction up to 3 years later. These studies converge on the finding that human‐coded redemption and automated linguistic analysis are complementary and nonredundant methods of analyzing life narratives, and considerations for the study of life narratives are discussed.     

Sentiment Analysis and Social Cognition Engine (SEANCE): An automatic tool for sentiment,social cognition,and social-order analysis

This study introduces the Sentiment Analysis and Cognition Engine (SEANCE), a freely available text analysis tool that is easy to use, works on most operating systems (Windows, Mac, Linux), is housed on a user’s hard drive (as compared to being accessed via an Internet interface), allows for batch processing of text files, includes negation and part-of-speech (POS) features, and reports on thousands of lexical categories and 20 component scores related to sentiment, social cognition, and social order. In the study, we validated SEANCE by investigating whether its indices and related component scores can be used to classify positive and negative reviews in two well-known sentiment analysis test corpora. We contrasted the results of SEANCE with those from Linguistic Inquiry and Word Count (LIWC), a similar tool that is popular in sentiment analysis, but is pay-to-use and does not include negation or POS features. The results demonstrated that both the SEANCE indices and component scores outperformed LIWC on the categorization tasks.