Bayesian inference for psychology. Part I: Theoretical advantages and practical ramifications

Bayesian parameter estimation and Bayesian hypothesis testing present attractive alternatives to classical inference using confidence intervals and p values. In part I of this series we outline ten prominent advantages of the Bayesian approach. Many of these advantages translate to concrete opportunities for pragmatic researchers. For instance, Bayesian hypothesis testing allows researchers to quantify evidence and monitor its progression as data come in, without needing to know the intention with which the data were collected. We end by countering several objections to Bayesian hypothesis testing. Part II of this series discusses JASP, a free and open source software program that makes it easy to conduct Bayesian estimation and testing for a range of popular statistical scenarios (Wagenmakers et al. this issue).     

Bound to Feel Bad About Oneself: Relations Between Attachment and the Self-conscious Emotions of Guilt and Shame in Children and Adolescents

Two cross-sectional studies were conducted to explore the relationship between attachment and the self-conscious emotions of guilt and shame in childhood. Study 1 was performed in non-clinical children aged 9–13 years (N = 688) who completed a single-item measure of attachment style and a vignette-based instrument for assessing guilt and shame. Results showed that children who classified themselves as insecurely attached displayed higher levels of shame and maladaptive types of guilt as compared to securely attached children. Study 2 was conducted in adolescents aged 12–18 years (N = 135), of whom the majority was referred to a clinical setting because of externalizing problems. Adolescents filled in a dimensional scale for measuring attachment quality to parents and peers and the above mentioned vignette-based instrument of guilt and shame. It was found that the clinical adolescents generally exhibited lower levels of self-conscious emotions as compared to non-clinical adolescents. Within this clinical group, communication to parents and peers was associated with higher levels of self-conscious emotions, and alienation was accompanied by higher levels of maladaptive forms of guilt and shame. Altogether, these results fit with the theory that attachment (in)security is involved in people’s proneness to experience self-conscious emotions.     

Bayesian tests of measurement invariance

Random item effects models provide a natural framework for the exploration of violations of measurement invariance without the need for anchor items. Within the random item effects modelling framework, Bayesian tests (Bayes factor, deviance information criterion) are proposed which enable multiple marginal invariance hypotheses to be tested simultaneously. The performance of the tests is evaluated with a simulation study which shows that the tests have high power and low Type I error rate. Data from the European Social Survey are used to test for measurement invariance of attitude towards immigrant items and to show that background information can be used to explain cross‐national variation in item functioning.     

Robustness of the cognitive gains in 7-month-old bilingual infants: A close multi-center replication of Kovács and Mehler (2009)

We present an exact replication of Experiment 2 from Kovács and Mehler's 2009 study, which showed that 7-month-old infants who are raised bilingually exhibit a cognitive advantage. In the experiment, a sound cue, following an AAB or ABB pattern, predicted the appearance of a visual stimulus on the screen. The stimulus appeared on one side of the screen for nine trials and then switched to the other side. In the original experiment, both mono- and bilingual infants anticipated where the visual stimulus would appear during pre-switch trials. However, during post-switch trials, only bilingual children anticipated that the stimulus would appear on the other side of the screen. The authors took this as evidence of a cognitive advantage. Using the exact same materials in combination with novel analysis techniques (Bayesian analyses, mixed effects modeling and cluster based permutation analyses), we assessed the robustness of these findings in four babylabs (N = 98). Our results did not replicate the original findings: although anticipatory looks increased slightly during post-switch trials for both groups, bilingual infants were not better switchers than monolingual infants. After the original experiment, we presented additional trials to examine whether infants associated sound patterns with cued locations, for which we did not find any evidence either. The results highlight the importance of multicenter replications and more fine-grained statistical analyses to better understand child development.

Highlights

• We carried out an exact replication across four baby labs of the high-impact study by Kovács and Mehler (2009).
• We did not replicate the findings of the original study, calling into question the robustness of the claim that bilingual infants have enhanced cognitive abilities.
• After the original experiment, we presented additional trials to examine whether infants correctly associated sound patterns with cued locations, for which we did not find any evidence.
• The use of novel analysis techniques (Bayesian analyses, mixed effects modeling and cluster based permutation analyses) allowed us to draw better-informed conclusions.