The study was approved by the local ethics committee (blinded to keep anonymity) and conducted as part of an overarching project on socio-emotional consequences of CM. The study was conducted in two steps: 1) creation of stimulus material (video recordings of target participants with and without a history of CM) and evaluation by confederates (members of the study team) during a first-acquaintance paradigm (FAP), involving an interaction between target participants and confederates; 2) evaluation of video recordings by independent raters (zero-acquaintance paradigm), involving no interaction between target participants and raters. All participants from both step 1 and step 2 gave written informed consent prior to participation.
2.1. Participants
Participants of step 1 (individuals with and without self-reported history of CM = target participants) were recruited via online social media platforms, flyers, mailing lists, from a study pool, and in collaboration with out-patient clinics in the area of (blinded to keep anonymity). Individuals aged 18–65 years with normal or corrected-to-normal vision that are native German speakers (or equivalent) were included. Exclusion criteria were antipsychotic, benzodiazepine, or tricyclic antidepressant medication, acute suicidality, lifetime psychotic symptoms, substance abuse or dependency (past 12 months), pregnancy, and physical health problems affecting psychophysiological measurements (that were conducted as part of the overarching project). The initial sample of step 1 consisted of almost twice as many target participants with CM (CM+) (n = 70) as without CM (CM-) experiences, as recruited through the main project. To match the two groups (i.e., to enable comparability between groups regarding gender, age, and education level), a random selection by matched subgroups was applied which resulted in a final sample of 40 target participants (26 female) in the CM + and 40 target participants (25 female) in the CM- group. Target participant’s characteristics are visualised in Table 1.
Table 1
Target’s Summary Statistics
|
CM+
(n = 40)
|
CM-
(n = 40)
|
|
|
n
|
%
|
n
|
%
|
Group comparison
|
Female gender
|
26
|
65.0
|
25
|
62.5
|
ns
|
Anxiety disorders c
|
10
|
25.0
|
4
|
10.0
|
ns
|
Obsessive-compulsive disorder c
|
4
|
10.0
|
1
|
2.5
|
ns
|
Affective disorders c
|
6
|
15.0
|
1
|
2.5
|
ns
|
Eating disorders c
|
0
|
0.0
|
1
|
2.5
|
ns
|
Sleeping disorders c
|
6
|
15.0
|
0
|
0.0
|
χ2 [1] = 6.49 a
|
BPD
|
6
|
15.0
|
0
|
0.0
|
χ2 [1] = 6.49 a
|
SSRI medication
|
6
|
15.0
|
1
|
2.5
|
ns
|
Other antidepressant medication
|
4
|
10.0
|
0
|
0.0
|
ns
|
|
M
|
SD
|
M
|
SD
|
|
Age (years)
|
33.00
|
13.69
|
32.85
|
12.46
|
ns
|
Educational group
|
2.46
|
0.82
|
2.4
|
0.78
|
ns
|
BDI-2
|
10.75
|
9.43
|
3.45
|
3.62
|
CM + > CM-; U = 353.0 b
|
CTQ emotional neglect
|
16.37
|
5.11
|
6.53
|
1.39
|
CM + > CM-; U = 060.5 b
|
CTQ physical neglect
|
8.35
|
3.06
|
5.18
|
0.39
|
CM + > CM-; U = 225.5 b
|
CTQ emotional abuse
|
12.00
|
5.34
|
5.50
|
0.85
|
CM + > CM-; U = 164.0 b
|
CTQ physical abuse
|
8.13
|
4.28
|
5.13
|
0.34
|
CM + > CM-; U = 435.0 b
|
CTQ sexual abuse
|
8.53
|
6.13
|
5.00
|
0.00
|
CM + > CM-; U = 460.0 b
|
PDS
|
2.28
|
2.11
|
0.98
|
1.27
|
CM + > CM-; U = 680.0 b
|
Notes: For count data comparison chi-square test and Fisher's exact test were used. Educational groups consisted of 4 levels: 1 = 11 years of education, 2 = 14–15 years of education, 3 = 19–22 years of education and 4 = 24–26 years of education. All p-values were computed two-sided. CM + = child maltreatment group, CM- = non child maltreatment group, ns = non-significant, a * = p < .05, b *** = p < .001, c assessed with Mini-DIPS, BPD = Borderline Personality Disorder, SSRI = Selective Serotonin Reuptake Inhibitor, BDI-2 = Beck Depression Inventory 2, CTQ = Childhood Trauma Questionnaire, PDS = Post-Traumatic Stress Diagnostic Scale |
The final sample of the confederates of step 1, consisted of 17 psychology student raters (11 female), with a mean age of 25.9 years, and a mean education level of 4.22 (representing the (blinded to keep anonymity) schooling system, see further description in section 2.2).
Participants (raters) of step 2 of the study were recruited via online social media platforms, mailing lists, and from a pool of former study participants. Individuals aged 18–65 years with normal or corrected-to-normal vision that are native German speakers (or equivalent) were included. The final sample consisted of 100 raters (67 female). Raters had a mean age of 28.8 years and a mean education level of 2.53 (representing the (blinded to keep anonymity) schooling system, see further description in section 2.2.)
2.2 Psychometric Assessment
CM was measured with the German version of the Childhood Trauma Questionnaire, short form (CTQ-SF; 28), German translation and validation of Bader et al., 2009 (29) to categorize target participants into CM + and CM- group in step 1. Internal consistency for the subscores is high (α > .81), except for the physical neglect subscale (α = 0.49). Nevertheless, the CTQ is a widely used measurement (30). In the current study, CM was rated as existent if a CTQ value was higher than “none / minimal” in at least one of the subscales according to Bernstein and colleagues (2003). The CM- group comprised individuals with a CTQ value of “none / minimal” on all subscales.
The following additional assessments of the target sample were conducted in step 1: 1) Depressive symptoms were measured using the German version of the Beck’s Depression Inventory 2 (BDI-II; 31). It is a self-report measure for the assessment of the severity of depressive symptoms over the past week and comprises 21 items, which can be added up to a sum score of 0–63, with a good validity and reliability (32). 2) Current mental disorders (affective disorders, obsessive-compulsive disorders, anxiety disorders, eating disorders, sleeping disorders) were assessed using the diagnostic interview for mental disorders Mini Diagnostisches Interview bei psychischen Störungen (Mini-DIPS; 33). The Mini-DIPS is a short, semi-structured clinical interview to assess the most common mental disorders (excluding personality disorders) according to the DSM-5. 3) The number of experienced trauma types was assessed using the trauma checklist of the Post-Traumatic Stress Diagnostic Scale (PDS; 34). This section of the instrument corresponds to stressor criterion A of the DSM-5 for PTSD and demonstrates excellent internal consistency and test–retest reliability, and good convergent validity with the PTSD Checklist - Specific Version and the PTSD Symptom Scale - Interview Version for DSM–5 (35). 4) For the assessment of BPD, the McLean screening instrument for borderline personality disorder (MIS-BPD; 36, 37) was utilized. This self-report measure is a screening instrument based on a subset of the questions that comprise the borderline module of the Diagnostic Interview for DSM-IV personality disorders, yielding both good sensitivity and specificity for the diagnosis of DSM-IV BPD (36). 5) Social interaction anxiety was measured with the Social Interaction Anxiety Scale (sias; 38), a self-report questionnaire assessing social interaction anxiety defined as “distress when meeting and talking with other people” and includes 20 items on a 5-point Likert-scale. It shows good reliability (retest-reliability: >.90; Cronbach’s alpha = .86) (38). 6) Social support was measured using the Fragebogen zur sozialen Unterstützung (F-SozU K22; 39). This self-report questionnaire assesses social support with 22 items and shows good reliability (Cronbach’s alpha = .81-.93) (39). 7) Rejection sensitivity was assessed with the Rejection Sensitivity Questionnaire (rsq; 40, 41), which is a self-report questionnaire assessing trait rejection sensitivity with 18 items with good reliability and validity (40).
Additionally, all participants’ educational levels were evaluated. Four categories were used; 1 = up to 13 years of education (mandatory school years), 2 = up to 18 years of education (high school degree), 3 = up to 23 years of education (university degree; Bachelor or higher) and 4 = more than 23 years of education (university degree; PhD or higher).
2.3. Material
2.3.1 Production of Video-Material and Zero-Acquaintance (Thin Slices) Paradigm
The video material comprised videos of 40 target participants of the CM + group and 40 target participants of the CM- group. All target participants performed the thin slices paradigm (TSP) (13, 42) while being filmed. In this paradigm, target participants were asked about their favourite meal, colour, hobby, book, movie, animal, past vacation, and holiday destination, while sitting in front of a white wall. Targets could freely decide whether they wanted to just name their answer to each category or provide further explanation. After the videos had been collected, sound and video track were separated from each other and videos cut at 30 seconds. In part 2, videos were presented to the independent raters without audio trace to exclude potential effects of speech content or prosody, based on the procedure by Hepp and colleagues (2018).
2.3.2 First-Acquaintance Paradigm
During the FAP, target participants held a short three-minute conversation with a same sex confederate via skype for business. The online interaction (rather than an in-person interaction) was chosen due to regulatory aspects (mandatory use of facemasks, which might have critically hampered the interpretation of facial emotion expression) of the Covid-19 pandemic. The three-minute interaction consisted of a standardized small-talk conversation. Target participants were told that the interaction partner was another study participant in order to create a close to real-life condition. Confederates had a set of questions and answers (e.g., “have you participated in a study before?”, “yes, this is my second participation”, “do you live in (blinded to keep anonymity)?” etc.) which they went through sequentially. If all questions had been asked, confederates initiated no more conversation. After three minutes, the experimenter broke off the dialogue. Directly after the interaction, we debriefed the target participants.
2.3.3 Trustworthiness, Likeability, and Cooperativeness Ratings
Raters of step 2 saw all 80 target videos and rated targets on likeability, trustworthiness, and cooperativeness on a 7-point Likert-scale. Similarity ratings were also collected on a 7-point Likert-scale. In an identical manner, confederates of the FAP rated the target participants.
To measure the target participant’s objective cooperativeness, the dictator game (DG; 27) was used. The dictator game is an economic game to assess cooperative behaviour. A fixed amount of money (here: 20 (blinded to keep anonymity) in 1 (blinded to keep anonymity) coins) can be divided between oneself and an unknown third person. Both parties are anonymous. The money is taken in private and participants are informed that someone unknown to them (i.e., not the person who serves as their experimenter) will open the envelope at the end of the participation.
2.3.4 FaceReader™
Objective measure of PA and NA display was assessed with the software FaceReader™ version 8 (43). For PA and NA display, the average percentage of positive or negative emotion detection (happy% + sad% + angry% + disgusted% + fearful%) during the TSP was assessed. Each video was calibrated manually and the sample rate was set to every second as suggested by the FaceReader™ manual 8 (44). The FaceReader™ is a valid measurement tool for emotional facial expressions, with 88% accuracy (43).
2.4. Procedure
The overarching project comprised two laboratory appointments. First, eligible targets were screened via telephone for inclusion and exclusion criteria and then scheduled an appointment for a first assessment in the laboratory, during which graduate psychology students trained and supervised by an experienced licensed psychotherapist (last author), assigned the questionnaires (sias, F-SozU K22, rsq) and conducted clinical interviews (CTQ-SF, BDI-II, MIS-BPD, Mini-Dips, PDS checklist). Target participants received a written study information and signed an informed consent form. The second laboratory visit comprised several emotion recognition paradigms (part of the overarching project not assessed for the current study), a personal space paradigm (not assessed for the current study; for further description see (blinded to keep anonymity) (under review), as well as the above described TSP, DG, and FAP conducted for part 1 of the current study. At the end of the second visit, we debriefed participants. They were reimbursed with 20 (blinded to keep anonymity) per hour for their participation in each study visit.
For part 2, eligible raters scheduled an appointment for the assessment in our laboratory. Each rater completed an informed consent form and then rated all 80 target videos, collected in step 1. At the end of the visit, participants received either course credits (1 credit per hour) or monetary compensation (20 (blinded to keep anonymity) per hour) for their participation.
2.5. Planned Statistical Analyses
Statistical analyses were calculated in R, version 4.2.1 (R Core Team, 20122). As pre-registered, it was planned to add similarity ratings to all models as a control variable. The first set of models to test Hypothesis 1 (individuals with CM experiences are evaluated as less likeable, trustworthy, and cooperative by independent raters at zero-acquaintance, compared to unexposed controls) comprised three separate regression analyses via lm function using the stats package for the influence of group allocation (CM + vs. CM-; predictor variable) on each criterion variable, i.e. average likeability, trustworthiness, and cooperativeness ratings by independent raters from part 2. The second set of models were related to Hypothesis 2 (individuals with CM experiences are evaluated as less likeable, trustworthy, and cooperative by confederates at first-acquaintance, compared to unexposed controls), planned to be tested by three separate regression analyses, with the ratings by confederates as criterion variables. As confederate ratings were missing for four participants (two of the CM- and two of the CM + group), a total of 76 ratings were collected. Attractiveness ratings were planned to be additionally added to models of confederate ratings as a control variable.
To test Hypothesis 3 (individuals with CM experiences express less PA and more NA compared to unexposed controls), two separate regression analyses via lm function for the influence of group (predictor variable) on PA and NA display (criterion variables) were intended to be conducted.
For exploratory analyses, a t-test with independent samples was conducted to test whether targets differed in their objective cooperativeness (as assessed with the dictator game). Furthermore, fifteen regression analyses were conducted to test whether the three rating dimensions (criterion variables) were related to self-reported levels of depressive symptoms, severity of CM, social anxiety, social support, and rejection sensitivity (all dimensional predictors).
Finally, the planned (according to pre-registration) mediation models between ratings and group, as well as overall levels of emotion expression, were not conducted, as no significant differences between study groups in ratings or affect display were found (see section results). Instead, exploratory (non-preregistered) analyses were conducted to better understand the unexpected findings and their deviation from previous research (8, 25, 26). More specifically, a possible influence of BPD (dimensional predictor) and complex PTSD symptoms (dimensional predictor) on each of the three rating dimensions (criterion variables), were assessed via lm function.