Using the LIWC program to study dreams.

Abstract

This article presents the results of an analysis of a large set of dream reports (N ϭ 5,208) using the Linguistic Inventory and Word Count (LIWC) system of Pennebaker, Boyd, Jordan, and Blackburn (2015). The findings indicate that, in comparison with other kinds of texts studied by LIWC, dream reports are distinctive in having high frequencies of the following language categories: focus on the past, first-person singular words, personal pronouns, authenticity, dictionary words, motion, space, and home. The dream reports have relatively low frequencies of these LIWC categories: informal language, focus on the present, assent, positive emotions, clout, second-person references, affective processes, and quotation marks. In addition, the LIWC analysis was able to identify and distinguish between the key content features of recent dreams, nightmares, and lucid dreams. These results confirm earlier findings of McNamara (2008) and Hawkins and Boyd (2017) and support the further use of LIWC in dream research, in coordination with other empirical methods of study.

Setting the pace: examining cognitive processing in MOOC discussion forums with automatic text analysis

“…Cognitive processing reflects a level of language complexity (Van Swol, Prahl, Kolb, Lewis, & Carlson, 2016) and further demonstrates the level to which the writer has organized their thoughts (Cohn, Mehl, & Pennebaker, 2004). This is supported by Bulkeley and Graves (2018) who assert that the higher the number of dictionary words used, the more use of formal language in the post.…”

“…LIWC has been used in various contexts, including predicting final course grades (Robinson, Navea, & Ickes, 2013), influence within groups and group dynamics (Van Swol & Kane, 2018), studies on deception (Newman, Pennebaker, Berry, & Richards, 2003), and interactions between people (Kacewicz, Pennebaker, Davis, Jeon, & Graesser, 2014). The current version of LIWC has about 100 different categories which focus on topics like grammar, vocabulary usage, perceptions, emotions, social processes, and personal concerns (Bulkeley & Graves, 2018). Fast et al (2017) assert that LIWC is now the standard-bearer for psychometrically validated categories.…”

“What Goes Up Must Come Down”—Elevators in a Long Dream Series

“…Simple dream content scales like coding the presence or absence of specific topics showed high interrater reliabilities (Schredl et al, 2004) and, thus, possible bias due to the self-coding should be minimal. On the other hand, it should be noted that the present approach allows more detailed analyses than using digital word search templates (Bulkeley, 2014(Bulkeley, , 2015(Bulkeley, , 2018Bulkeley & Graves, 2018) because automatized approaches would, for example, not allow to associate the dream anxiety to using the elevator. So, future studies might apply both approaches, that is, doing a word search with subsequent coding by external judges of the hits in order to keep the workload manageable.…”

A Corpus Based Study of Addiction Treatment Manuals

“…TSF is closely aligned with AA, which was written in the 1930s when public discourse was less open than today, a trend described in a Pennebaker interview (Sutton, 2017). Bulkeley and Graves (2018) point out that in LIWC2015, the more a writer filters what is being said for their audience, the lower the authentic score and vice versa. Between the three explanations, the third is most likely because TSF has its basis in AA, whose literature was written long before the other manuals, during an era of less openness and more filtering.…”