Are sad children more believable? A systematic review of the relationship between emotional demeanour of child victims and juror credibility judgements

Figures & data

Table 1. PICO framework table and inclusion and exclusion criteria.

Study characteristic	Inclusion criteria	Exclusion criteria	Rationale
Population.	Jury eligible adults.	Jury ineligible adults.	To replicate the jury eligible population in real trials.
Interventions.	The child’s emotional presentation (e.g. sad, happy, angry, neutral) is experimentally manipulated between subjects.	Other methods, such as field research and qualitative studies.	To focus the results of the review.
Comparators.	Different emotional presentations, including sadness, anger, positive or neutral.	Other types of emotional presentation.	To focus the results of the review.
Outcome.	Judgments of child credibility and ratings of defendant guilt made by mock jurors.		To answer the research question.
Age of child victims used in stimulus.	4–18 years old.	Victims younger than 4 and older than 18.	A child aged under 4 may have less developed linguistic skills (e.g. La Rooy et al., 2016). A victim aged over 18 is considered an adult in most countries globally.
Timeframes.	Studies published within the last 20 years (1998–2019).	Studies conducted prior to 1998.	To ensure the included studies are current.
Publication.	Published studies only.	Unpublished studies.	Scoping exercise provided no hits for unpublished literature and the criminal justice system favours published research (e.g. Daubert Criteria regarding admissibility of evidence).
Language.	English language.	Non-English language.	International research more likely to be published in English language only.

Figure 1. Prisma flowchart of article screening processes.

Table 2. Quality assessment questions on the CASP (Critical Appraisal Skills Programme, 2019).

Number	Question
1	Did the study address a clearly focused issue?
2	Was the assignment of participants to conditions randomised?
3	Were all of the participants who entered the trial properly accounted for at its conclusion?
4	Were participants and researchers ‘blind’ to treatment?
5	Were the groups similar at the start of the trial?
6	Aside from the experimental intervention, were the groups treated equally?
7	How large was the outcome measure?
8	How precise was the estimate of the outcome?
9	Can the results be applied to the local population, or in your context?
10	Were all clinically important outcomes considered?
11	Are the benefits worth the harms and costs?

Table 3. Extracted article information.

Author and year	Participants	Study aim	Child age and gender	Crime type	Measures	Experimental manipulation	Quality score
Bederian-Gardner et al. (2017)	Undergraduate students who received course credit (N = 354. 62.4% female, 37.6 male. M age = 20.5 years, range 18–40 years).	To explore the role of empathy in adult judgements of child victims.	5 or 13 years Male or female	Sexual assault	-Participants completed the Child Victim Empathy Scale pre and post-test (validated scale). -Rated how sad the child appeared (7-point Likert scale). -Rated how believable the child was (6-point Likert scale). -Rendered a decision regarding defendant guilt (dichotomous choice) and confidence in that decision (12-point Likert scale).	-Emotional presentation of low, medium, and high sadness (photo).	94.5%
Cooper et al. (2014)	Sample one: Undergraduate students who received course credit (N = 308. 51.9% male, 48.1% female. M age = 20.8 years, range 18–47). Sample two: Jurors released from duty paid $10 (n=267. 49.1% male, 50.9% female, M age = not stated, range 20–80). (total N=575).	To explore if older jurors, with more experience of children, would rate an emotional child as more credible than a younger juror with less experience of children.	6 or 12 years Male or female	Sexual assault	-Participants completed a questionnaire measuring child and defendant intelligence, honesty, accuracy, believability, confidence, likeability, consistency and memory for understanding of the event (6-point Likert scales). Questions were modelled from previous studies. -Rendered a decision regarding defendant guilt (dichotomous choice) and a guilty rating question were asked (the latter a 6-point Likert scale). Final question asked how emotional the child was (6-point Likert scale).	-Emotional presentation of child as sad and tearful versus calm and neutral (line drawing).	76.5%
Golding et al. (2003)	Undergraduate students who received course credit (N = 150. 54.7% female, 45.3% male. No information on age presented apart from over the age of 18).	To explore the effect of child victim demeanour on jury credibility ratings in a sexual assault case.	6 or 15 years Female	Sexual assault	-Participants made decisions regarding guilt (dichotomous choice) and ratings of believability of child victims (10-point Likert scale). If guilty decision given, participants were asked to give a sentencing opinion (based on the local laws for the area; up to life for the 6 year old child and up to 20 years for the 15 year old). -Confidence in the verdict was measured (10-point Likert scale).	- Emotional presentation of either calm, teary, or hysterical crying (trial summary description and courtroom drawing of the child).	90%
Landström et al. (2015)	Undergraduate students who received a €10 cinema ticket (N = 155. 58% female, 42% male. M age = 23.21 years, range 20–38 years).	To explore the impact of child victim demeanour during testimony and the effects of the camera perspective on credibility judgements.	8 years Male or female	Harassment	-Participants completed rating scales for authenticity, credibility and expectancy confirmation (7-point Likert scales).	- Emotional presentation of either sad or neutral (video recording).	81%
Melinder et al. (2016)	Undergraduate students (N = 465. 66% female, 44% male. M = 23.43 years. Range 15–64 years).	To explore the impact of child victim demeanour during testimony and the effects of presentation mode on credibility judgements.	11 or 13 years Female	Physical abuse	-Participants rated the credibility and believability of the child victim and the reliability of the testimony provided (7-point Likert scales).	- Four different emotional presentations of sad, angry, happy or neutral. - Video recording, audio recording, or written transcripts to examine presentation mode.	81%
Wessel et al. (2013)	Sample one: Undergraduate students (n = 162. 63% female, 37% male. M age = 28 years, range not stated). Sample two: Child protection workers (n = 154. 89% female. M age = 39 years, range not stated).	To explore the impact of child victim demeanour during testimony and the impact of experience of working with children on credibility judgements.	11 years old Female	Physical abuse	-Participants rated credibility of the child (7-point Likert scale) and likelihood of guilt (percentage scale from 1 to 100).	- Four different emotional presentations of sad, angry, happy or neutral. (Video recording).	72%
Wessel et al. (2016)	Undergraduate students (N = 119. 59% female, 41% male. M age = 22.7 years, range 19–49 years).	To explore the impact of child victim demeanour during testimony on credibility judgements.	11 years Female	Sexual abuse	-Participants rated credibility (7-point Likert scale) and rated the likelihood of the perpetrator being guilty (100-point percentage scale).	- Four different emotional presentations of sad, angry, happy, or neutral. (Video recording).	76.5%
Regan and Baker (1998)	Undergraduate students (N = 31. 64.5% female, 35.5% male. M age = 18.39 years, range not stated) who received course credit.	To explore the impact of child victim demeanour during testimony on credibility judgements.	6 years Female	Sexual abuse	-Participants rated the credibility, honesty and reliability of the child and the likelihood of guilt (9-point Likert scales?).	- Emotional presentation of the child victim (crying versus neutral presentation).	81%

Table 4. Main findings of studies.

Author and year	Findings
Bederian-Gardner et al. (2017)	- Increased appraised sadness significantly predicted increased perceived believability. - Participants who rated the child as believable were more likely to give guilty verdicts.
Cooper et al. (2014)	- Overall, students gave more guilty verdicts compared to jurors, and female students were more likely to give children higher credibility ratings than male students. Amongst jurors, females were more likely to render guilty verdicts than male jurors, and females were more likely to rate younger children as more credible than older children. - Participants who rated the child as more emotional were more likely to give guilty verdicts and considered both the older and younger child as more credible than participants who rated the child as less emotional.
Golding et al. (2003)	- The child presenting as ‘teary’ was more likely to be believed and receive more guilty verdicts than the child presenting as ‘calm’ or ‘hysterically crying’. - No significant difference in the mean number of guilty verdicts between the calm or hysterical conditions.
Landström et al. (2015)	- Participants who observed a sad emotional child were more likely to believe that the child had experienced the abuse, compared to the child in the neutrally presenting condition. - Compared to the neutrally presenting child, participants rated the sad child as making a more credible impression and rated that the child’s demeanour better matched their expectations.
Melinder et al. (2016)	- Child victims were rated as more credible when they displayed the sad emotion compared with the angry and positive emotional expressions. - Written presentation mode gained higher credibility ratings than audio or visual recordings.
Wessel et al. (2013)	- Two groups (lay vs CPS) correlated highly on their ratings of credibility and subsequent probability of guilt. - Lay participants overall rated the victims’ credibility significantly lower and with less guilty decisions. - Both sample groups rated the sad condition as more credible than the angry, happy and neutral presentations. This demonstrates that CPS workers are governed by the same social norms as lay people. - Highest ratings of credibility and guilty decisions were in the sad condition, followed by the neutral, then the anger condition, and were lowest for the positive condition.
Wessel et al. (2016)	- No significant difference in credibility ratings between the sad and neutral condition. Therefore, these two conditions were combined and considered under broader term of ‘emotional valence’. - Emotional valence was shown to elicit higher credibility ratings than the angry or positive conditions. - Participants who rated the child has more credible were more likely to give guilty verdicts.
Regan and Baker (1998)	- Participants who read about a child who cries upon confronting the defendant were more likely than those who read about a calm child, to perceive the child victims as honest, credible, accurate and reliable. - Participants who read a child was crying were more likely to believe the victim had been abused and conclude that the defendant was guilty.

La Rooy, D., Heydon, G., Korkman, J., Myklebust, T. (2016). Interviewing child witnesses. In G. Oxburgh, T. Myklebust, T. Grant, & R. Milne (Eds.), Communication in investigative and legal contexts: Integrated approaches from forensic psychology, linguistics and law enforcement (1st ed., pp. 55–79). John Wiley & Sons, Ltd. https://doi.org/10.1002/9781118769133.ch4

 Google Scholar

Critical Appraisal Skills Programme. (2019). CASP checklist for randomised control trials. https://casp-uk.net

 Google Scholar

**Bederian-Gardner, D., Goldfarb, D., & Goodman, G. S. (2017). Empathy's relation to appraisal of the emotional child victims. Applied Cognitive Psychology, 31(5), 488–499. https://doi.org/10.1002/acp.3345

 Web of Science ®Google Scholar

**Cooper, A., Quas, J. A., & Cleveland, K. C. (2014). The emotional child victims: Effects on juror decision-making. Behavioral Sciences and the Law, 32(6), 813–828. https://doi.org/10.1002/bsl.2153

 PubMed Web of Science ®Google Scholar

**Golding, J. M., Fryman, H. M., Marsil, D. F., & Yozwiak, J. A. (2003). Big girls don't cry: The effect of child victims demeanour on juror decisions in a child sexual abuse trial. Child Abuse and Neglect, 27(11), 1311–1321. https://doi.org/10.1016/j.chiabu.2003.03.001

 PubMedGoogle Scholar

Landström, S., Ask, K., Sommar, C., & Willén, R. (2015). Children's testimony and the emotional victim effect. Legal and Criminological Psychology, 20(2), 365–383. https://doi.org/10.1111/lcrp.12036

 Web of Science ®Google Scholar

**Melinder, A., Burrell, L., Eriksen, M. O., Magnussen, S., & Wessel, E. (2016). The emotional child victims effect survives presentation mode. Behavioral Sciences and the Law, 34(1), 113–125. https://doi.org/10.1002/bsl.2232

 PubMed Web of Science ®Google Scholar

**Wessel, E., Magnussen, S., & Melinder, A. M. D. (2013). Expressed emotions and perceived credibility of child mock victims disclosing physical abuse. Applied Cognitive Psychology, 27(5), 611–616. https://doi.org/10.1002/acp.2935

 Web of Science ®Google Scholar

**Wessel, E. M., Eilertsen, D. E., Langnes, E., Magnussen, S., & Melinder, A. (2016). Disclosure of child sexual abuse: Expressed emotions and credibility judgments of a child mock victim. Psychology, Crime and Law, 22(4), 331–343. https://doi.org/10.1080/1068316X.2015.1109087

 Web of Science ®Google Scholar

**Regan, P., & Baker, S. (1998). The impact of child victims demeanour on perceived credibility and trial outcome in sexual abuse cases. Journal of Family Violence, 13(2), 187–195. https://doi.org/10.1023/A:1022845724226

 Google Scholar