Abstract

Objective: The Reflective Function (RF) Scale is the original validated measure of mentalizing, applied to Adult Attachment Interview (AAI) transcripts. This study aimed to replicate the preliminary validation of Computerized RF (CRF), a computerized text analysis measure of RF.

Methods: 49 women with borderline personality disorder (BPD; n = 20) and without (n = 29) completed diagnostic interviews and the AAI. AAIs were scored by certified coders using the RF Scale, and by computerized text analysis software using CRF dictionaries.

Results: CRF significantly correlated with RF. The frequency of High CRF words had moderate-to-strong correlations with RF in the total sample (r_s = .47), BPD group (r_s = .61), and non-BPD group (r_s = .41). Adding the frequency of Low CRF words to create a proportion of High and Low CRF words, and applying CRF to AAIs trimmed to include only items that “demand” RF, both produced significant, but not stronger, correlations with RF.

Conclusion: CRF is an efficient alternative to the RF Scale that can estimate mentalization. Near average levels, neither RF nor CRF significantly differentiated between groups, indicating mentalizing deficits may not always be present, detectable, or defining in BPD samples. While more research is needed, CRF is a valid assessment of RF when codable text is available.

摘要

目的：反射功能（RF）量表是用于成人依恋访谈（AAI）成绩单的经过验证的心理测度的原始量度。这项研究旨在复制计算机化射频（CRF）的初步验证，这是一种计算机化的射频文本分析方法。

方法： 49名患有边缘型人格障碍（BPD；n  = 20）且没有（n  = 29）的女性完成了诊断性访谈和AAI。由认证编码员使用RF量表对AAI进行评分，并使用CRF词典通过计算机文本分析软件对AAI进行评分。

结果： CRF与RF显着相关。高CRF词的频率与总样本中的RF（r _s  = 0.47），BPD组（r _s  = 0.61）和非BPD组（r _s  = 0.41）具有中等至强的相关性。添加低CRF词的频率以创建高CRF词和低CRF词的比例，并将CRF应用于修剪的AAI，以仅包括“需求” RF的项目，两者均与RF产生了显着但不强的相关性。

结论： CRF是可以估计心理状态的RF量表的有效替代方法。在接近平均水平的情况下，RF和CRF在组之间均未明显区分，表明在BPD样本中可能并非总是存在，可检测到或无法定义出心理缺陷。尽管需要进行更多的研究，但是当可编码文本可用时，CRF是对RF的有效评估。