Writing Trauma Narratives Increases Temporal Organization & Habituation

Emotional processing theory posits that the therapeutic benefits of exposure include habituation to the distress triggered by trauma memories (Lang, 1977) as well as their reorganization (Foa, Molnar, & Cashman, 1995). Trauma narratives among people with PTSD have been observed to have greater sensory, perceptual, and emotional references and interrupted temporal, causal and logical connections (O’Kearney & Perrott, 2006). Habituation across exposures is cited as an indicator of emotional processing (Foa & Kozak, 1986), but literature on linguistic changes to trauma narratives as indicators of emotional processing is more sparse.

Howard University Clinical Psychology PhD student Mary Katherine Howell examined changes in temporal references and subjective distress across written trauma narrative exposure sessions in persons with PTSD. She presented her findings in a poster presentation at the 2015 APA Convention and was selected a Student Poster Award winner for SCP’s Student Poster Award Competition.

This project, supported by an NIMH grant awarded to Principal Investigator and Mary Katherine’s mentor, Dr. Thomas A. Mellman, recruited thirty non-treatment seeking trauma-exposed urban residing African American participants (average age = 27, 43% male) from the greater DC area–27 met current DSM-IV PTSD diagnostic criteria and 3 met subthreshold PTSD criteria (i.e., two of the three symptom clusters, re-experiencing, avoidance, and  hyperarousal).

Participants completed four 30-minute writing sessions and were instructed to write about the traumatic event in as much detail as possible. Writing sessions 1 and 2 were 12 hours apart and sessions 3 and 4 were performed 1 week later, also 12 hours apart. Subjective Units of Distress (SUD) were assessed on a 0-100 scale 4 times each session. Habituation across sessions was operationalized as decreasing maximum Subjective Units of Distress (SUD) scores between sessions, and PTSD symptom severity was assessed at baseline and 1 week after sessions 2 and 4 with the Clinician Administered PTSD Scale (CAPS; Blake, et al., 1995). Narratives were analyzed with the Linguistic Inquiry and Word Count 2007 (LIWC2007; Pennebaker, Booth, & Francis, 2007) coding program for rates of temporal word usage. Temporal organization words were operationalized with the “time” variable of the LIWC, which includes words such as “end,” “until,” and “season.” Changes in narrative features and subjective distress served as indicators of between session emotional processing.

All participants who completed the protocol had reduced PTSD severity from baseline. Average reduction 33.667 CAPS score points (SD = 17.850), t (20) = 8.643, p < .001. A one-way repeated measures ANOVA revealed a significant effect of writing sessions on “time” variable words, F (3, 51) = 3.125, p < .05, η2= 0.115. Post-hoc comparison using Bonferroni revealed that the occurrence of time words significantly increased from writing session 2 to 4 (6.2% to 7.1%, p < .05) and trended towards significance from session 2 to 3 (6.2% to 7.4%, p = .052) . See Figure 1 below for a graph of the estimated marginal means of time words over writing sessions, and see Table 1 below for qualitative examples of the increase in time-related words across writing sessions

Figure 1. Estimated Marginal Means of Time Words over Writing Sessions


Table 1. Qualitative Examples from the Written Trauma Narratives

Qualitative Examples from the Written Trauma Narratives


A one-way repeated measures ANOVA revealed a significant effect of writing sessions on maximum SUD score, F (3, 60) = 6.727, p < .005, η2= 0.252. Post-hoc comparison using Bonferroni revealed that the maximum SUD score significantly decreased from writing session 1 to 4 (40 to 22, p < .05). Moreover, an  increase in time words from session 1 to 3 was strongly correlated with a decrease in maximum SUD score from session 2 to 3, r (21) = -0.465, p < 0.05. See Figure 2 below for a graph of the estimated marginal means of maximum subjective distress over writing sessions.

Figure 2. Estimated Marginal Means of Maximum Subjective Distress over Writing Sessions


The increase in the use of temporal organization-related words over the course of writing sessions suggest increased awareness of temporal relationships and ability to describe and organize details of the event. These findings indicate increased awareness of temporal relationships preceding and in association with habituation. Notably, the increase in time words was not linear across sessions. One possible explanation for the increase between sessions 2 and 3 and between sessions 2 and 4 is that participants became fatigued writing twice within the same 12 hours (sessions 1 and 2; sessions 3 and 4) and wrote narratives with less specific detail overall, but in the intervening week, the participant had enough time to process the effect of writing, thus, it had the predicted memory consolidation effect and increase in temporal words. Overall, signs of emotional processing appeared more robust over 1 week than between sessions that were 12 hours apart.

Trauma narrative writing as an intervention may be an invaluable tool for emotionally processing traumatic events and organizing and consolidating the traumatic memory. Future research should explore other LIWC variables on a larger population.

Author Bio

howell_headshotMary Katherine Howell obtained her B.A. in Psychology and Middle East Studies from Northwestern University and is currently a graduate student in the Clinical Psychology PhD Program at Howard University in Washington, DC. Her primary research interests are in predictive patterns of distress activation and habituation in PTSD as well as post-traumatic distress presentation in various cultural contexts, such as in urban settings, veteran populations, and Arab populations.

Discussion Questions

  1. How might clinicians use time references as markers of emotional processing and habituation during exposure-based therapies or other therapeutic modalities?
  2. What other linguistic features of trauma narratives might you expect to change with repeated writing sessions? Any other categories of words besides time-related words? Any grammatical features?


Blake, D.D., Weathers, F.W., Nagy, L.M., Kaloupek, D.G., Gusman, F.D., Charney, D.S., & Keane, T.M. (1995). The development of a clinician-administered PTSD scale. Journal of Traumatic Stress, 8(1), 75-90.

Foa, E.B. & Kozak, M.J. (1986). Emotional processing of fear: Exposure to corrective information. Psychology Bulletin, 99(1), 20-35.

Foa, E.B., Molnar, C., & Cashman, L. (1995). Change in rape narratives during exposure therapy for posttraumatic stress disorder. Journal of Traumatic Stress, 8(4), 675-690.

Lang P.J. (1977). Imagery in therapy: An information processing analysis of fear. Behavior Therapy, 8, 862-886.

O’Kearney, R., & Perrott, K. (2006). Trauma narratives in posttraumatic stress disorder: A review. Journal of Traumatic Stress, 19(1), 81-93.

Pennebaker, J.W., Booth, R.J., & Francis, M.E. (2007). LIWC2007: Linguistic inquiry and word count. Austin, Texas: liwc.net