Subsequent to transcription and encoding of all data, computer-aided content analysis proceeded employing the Linguistic Inquiry Word Count (LIWC) software, 2007 version, developed by Pennebaker et al. (6-10). Like previously-popular, if dated, alternatives in the computer assisted qualitative data analysis software (CAQDAS) field – Nudist and Atlas/ti (Barry 1-2) – LIWC analyzes and classifies natural text word-by-word, creates categorical frequencies, outputs measurements by word forms (nouns/pronouns, articles), psychological dimensions (affective, cognitive, social), relativity (time, space), and, though not applicable in this case, personal information (occupation, hobbies). CAQDAS software like LIWC is immensely helpful for coding large response sets even from the limited sample size of this study and thereby improves reliability of the highly subjective coding process.
Given the subtlety and likely interactions amongst many linguistic cues, nonetheless, manual coding will likely be necessary to augment results provided by LIWC. At this stage of data analysis, coding frameworks shall include:
- The balance between objective reporting and subjective attribution can be tested via contextual embedding, references to place and time that appear more frequently in objectively true statements than in those embellished with fabrication or attribution (Vrij 6).
- Criteria-Based Content Analysis (CBCA, Steller & Köhnken 221) which scans transcribed data for the presence of 19 criteria, among others: logical structure of statements, descriptions of interactions, accounts of affect or feelings experienced, unprompted corrections, and the like.
- Reality Monitoring (RM), which is based on the premise that ‘true-to-life’ accounts differ substantively from those rationalised or fabricated. The latter are more heavily imbued with internal or personal contexts and cognitive operations. RM has the stronger theoretical foundation and is more efficient and reliable than CNCA (Masip et al. 101-2; Sporer 376).
Further data analysis will comprise at least three distinct stages. The first shall consist of standard descriptive statistics to profile the categories created by LIWC, refinements created by manual coding, and the six scale items.
The second and third stages shall be necessitated by the requirement for cross-analysis between the open-ended item results on one hand and the six Likert-scale-type self-rating items on the other. Since the top-level analysis shall relate the scale results to internal versus external attributions, clearly a nominal variable, the Kruskal-Wallis H test seems appropriate for finding meaningful relationships across both data sets. This test is the non-parametric equivalent of the ANOVA when the assumptions about equal sub-sample sizes, independent and randomly-drawn samples, equal variances, and normal distribution of the data cannot be met.
Lastly, tests for differences of central tendency across categorical sub-groups can be conducted with the Mann–Whitney–Wilcoxon test, a non-parametric test for assessing whether two independent samples of observations come from the same distribution. This is the non-parametric alternative to Student’s t test, solving for significance of differences among medians and rank sums rather than means because the assumptions of random sampling, equal sub-samples and equality of distributions cannot be met within the parameters of the University of Nottingham’s rugby 1st XV.
Barry, Christine A. (1998) “Choosing Qualitative Data Analysis Software: Atlas/ti and Nudist Compared.” Sociological Research Online, 3 (3).
Masip, J. et al. “The Detection of Deception with the Reality Monitoring approach: A Review of the Empirical Evidence. Psychology, Crime, & Law (2005) 11, 99–122.
Pennebaker, J.W., R.J. Booth and M. E. Francis. Linguistic Inquiry and Word Count: LIWC 2007. Austin, TX: LIWC (2007).
Sporer, S. L. “The Less Traveled Road to Truth: Verbal Cues in Deception Detection in Accounts of Fabricated and Self-Experienced Events. Applied Cognitive Psychology (1997) 11, 373–397.
Steller, M., & G. Köhnken. “Criteria-based Content Analysis.” In D. C. Raskin (Ed.), Psychological methods in criminal investigation and evidence. New York: Springer-Verlag, 1989.
Vrij, Aldert. “Criteria-Based Content Analysis: A Qualitative Review of the First 37 Studies.” Psychology, Public Policy, and Law (2005) 11, 3–41.
Vrij, Aldert et al. “Cues to Deception and Ability to Detect Lies as a Function of Police Interview Styles.” Law Hum Behav (2007) 31: 499–518.