Trimbos Institute, Knowledge Centre of Mental Health and Addiction, Da Costakade 45, Post Office Box 727, 3500 Utrecht, The Netherlands.
Department of Social Sciences, Utrecht University, The Netherlands.
J Affect Disord. 2014 May;160:10-3. doi: 10.1016/j.jad.2014.01.005. Epub 2014 Jan 22.
The growing number of web-based psychological treatments, based on textual communication, generates a wealth of data that can contribute to knowledge of online and face-to-face treatments. We investigated whether clients' language use predicted treatment outcomes and adherence in Master Your Mood (MYM), an online group course for young adults with depressive symptoms.
Among 234 participants from a randomised controlled trial of MYM, we tested whether their word use on course application forms predicted baseline levels of depression, anxiety and mastery, or subsequent treatment adherence. We then analysed chat session transcripts of course completers (n=67) to investigate whether word use changes predicted changes in treatment outcomes.
Depression improvement was predicted by increasing use of 'discrepancy words' during treatment (e.g. should). At baseline, more discrepancy words predicted higher mastery level. Adherence was predicted by more words used at application, more social words and fewer discrepancy words.
Many variables were included, increasing the chance of coincidental results. This risk was constrained by examining only those word categories that have been investigated in relation to depression or adherence.
This is the first study to link word use during treatment to outcomes of treatment that has proven to be effective in an RCT. The results suggest that paying attention to the length of problem articulation at application and to 'discrepancy words' may be wise, as these seem to be psychological markers. To expand knowledge of word use as psychological marker, research on web-based treatment should include text analysis.
基于文本交流的网络心理治疗数量不断增加，产生了大量数据，这些数据可以为在线和面对面治疗提供知识。我们研究了客户的语言使用是否可以预测在线团体课程 Master Your Mood（MYY）中年轻抑郁患者的治疗结果和治疗依从性。
在 MYY 的一项随机对照试验中，我们对 234 名参与者进行了测试，检验他们在课程申请表上的用词是否可以预测抑郁、焦虑和掌握程度的基线水平，或随后的治疗依从性。然后，我们分析了完成课程的聊天会话记录（n=67），以调查用词变化是否可以预测治疗结果的变化。
治疗过程中“差异词”（如“应该”）的使用增加预示着抑郁的改善。在基线时，更多的差异词预示着更高的掌握程度。依从性可由申请时使用的更多词汇、更多的社交词汇和更少的差异词汇来预测。
纳入了许多变量，增加了巧合结果的机会。通过仅检查与抑郁或依从性有关的那些词汇类别，可以限制这种风险。
这是第一个将治疗过程中的用词与经 RCT 证明有效的治疗结果联系起来的研究。结果表明，关注申请时问题表达的长度和“差异词”可能是明智的，因为这些词似乎是心理标志物。为了扩大用词作为心理标志物的知识，应将基于网络的治疗研究纳入文本分析。
