A deep learning method for comparing Bayesian hierarchical models.

Bayesian model comparison (BMC) offers a principled approach to assessing the relative merits of competing computational models and propagating uncertainty into model selection decisions. However, BMC is often intractable for the popular class of hierarchical models due to their high-dimensional nested parameter structure. To address this intractability, we propose a deep learning method for performing BMC on any set of hierarchical models which can be instantiated as probabilistic programs. Since our method enables amortized inference, it allows efficient re-estimation of posterior model probabilities and fast performance validation prior to any real-data application. In a series of extensive validation studies, we benchmark the performance of our method against the state-of-the-art bridge sampling method and demonstrate excellent amortized inference across all BMC settings. We then showcase our method by comparing four hierarchical evidence accumulation models that have previously been deemed intractable for BMC due to partly implicit likelihoods. Additionally, we demonstrate how transfer learning can be leveraged to enhance training efficiency. We provide reproducible code for all analyses and an open-source implementation of our method. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Post-hypnotic suggestion improves confidence and speed of memory access with long-lasting effects.

In our study, we use the post-hypnotic suggestion of easy remembering to improve memory with long-lasting effects. We tested 24 highly suggestible participants in an online study. Participants learned word lists and recalled them later in a recognition memory task. At the beginning of the study, participants were hypnotized and the post-hypnotic suggestion to remember easily was associated with a cue that participants used during the recognition memory task. In a control condition, the same participants used a neutral cue. One week later, participants repeated both conditions with new word lists. Participants were significantly faster and more confident in their recognition ratings in the easy-remembering condition compared to the control condition, and this effect persisted over one week. Crucially, the increased speed and confidence in the easy-remembering condition did not affect memory accuracy. That makes our hypnosis intervention promising for patients experiencing subjective memory impairments. APA PSYCINFO CODES: 2343 (Learning and Memory), 2380 (Consciousness States), 3351 (Clinical Hypnosis).

Waldian t tests: Sequential Bayesian t tests with controlled error probabilities.

Bayesian t tests have become increasingly popular alternatives to null-hypothesis significance testing (NHST) in psychological research. In contrast to NHST, they allow for the quantification of evidence in favor of the null hypothesis and for optional stopping. A major drawback of Bayesian t tests, however, is that error probabilities of statistical decisions remain uncontrolled. Previous approaches in the literature to remedy this problem require time-consuming simulations to calibrate decision thresholds. In this article, we propose a sequential probability ratio test that combines Bayesian t tests with simple decision criteria developed by Abraham Wald in 1947. We discuss this sequential procedure, which we call Waldian t test, in the context of three recently proposed specifications of Bayesian t tests. Waldian t tests preserve the key idea of Bayesian t tests by assuming a distribution for the effect size under the alternative hypothesis. At the same time, they control expected frequentist error probabilities, with the nominal Type I and Type II error probabilities serving as upper bounds to the actual expected error rates under the specified statistical models. Thus, Waldian t tests are fully justified from both a Bayesian and a frequentist point of view. We highlight the relationship between Bayesian and frequentist error probabilities and critically discuss the implications of conventional stopping criteria for sequential Bayesian t tests. Finally, we provide a user-friendly web application that implements the proposed procedure for interested researchers. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Improving the efficiency of surveys with randomized response models: A sequential approach based on curtailed sampling.

Randomized response models (RRMs) aim at increasing the validity of measuring sensitive attributes by eliciting more honest responses through anonymity protection of respondents. This anonymity protection is achieved by implementing randomization in the questioning procedure. On the other hand, this randomization increases the sampling variance and, therefore, increases sample size requirements. The present work aims at countering this drawback by combining RRMs with curtailed sampling, a sequential sampling design in which sampling is terminated as soon as sufficient information to decide on a hypothesis is collected. In contrast to nontruncated sequential designs, the curtailed sampling plan includes the definition of a maximum sample size and subsequent prevalence estimation is easy to conduct. Using this approach, resources can be saved such that the application of RRMs becomes more feasible. An R Shiny web application is provided for simplified application of the proposed procedures. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

On the efficiency of the independent segments procedure: A direct comparison with sequential probability ratio tests.

In this comment, we report a simulation study that assesses error rates and average sample sizes required to reach a statistical decision for two sequential procedures, the sequential probability ratio test (SPRT) originally proposed by Wald (1947) and the independent segments procedure (ISP) recently suggested by Miller and Ulrich (2020). Following Miller and Ulrich (2020), we use sequential one-tailed t tests as examples. In line with the optimal efficiency properties of the SPRT already proven by Wald and Wolfowitz (1948), the SPRT outperformed the ISP in terms of efficiency without compromising error probability control. The efficiency gain in terms of sample size reduction achieved with the SPRT t test relative to the ISP may be as high as 25%. We thus recommend the SPRT as a default sequential testing procedure especially for detecting small or medium hypothesized effect sizes under H1 whenever a priori knowledge of the maximum sample size is not crucial. If a priori control of the maximum sample size is mandatory, however, the ISP is a very useful addition to the sequential testing literature. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

The truth revisited: Bayesian analysis of individual differences in the truth effect.

The repetition-induced truth effect refers to a phenomenon where people rate repeated statements as more likely true than novel statements. In this paper, we document qualitative individual differences in the effect. While the overwhelming majority of participants display the usual positive truth effect, a minority are the opposite-they reliably discount the validity of repeated statements, what we refer to as negative truth effect. We examine eight truth-effect data sets where individual-level data are curated. These sets are composed of 1105 individuals performing 38,904 judgments. Through Bayes factor model comparison, we show that reliable negative truth effects occur in five of the eight data sets. The negative truth effect is informative because it seems unreasonable that the mechanisms mediating the positive truth effect are the same that lead to a discounting of repeated statements' validity. Moreover, the presence of qualitative differences motivates a different type of analysis of individual differences based on ordinal (i.e., Which sign does the effect have?) rather than metric measures. To our knowledge, this paper reports the first such reliable qualitative differences in a cognitive task.

Predictors of focal to bilateral tonic-clonic seizures during long-term video-EEG monitoring.

OBJECTIVE: To determine predictors of focal to bilateral tonic-clonic seizures (FBTCS) during video-electroencephalography (EEG) monitoring (VEM). METHODS: All adult patients undergoing presurgical VEM from 2014 to 2015 in the department of epileptology were eligible (N = 229). Those with refractory focal epilepsy and epileptic seizures recorded during VEM were analyzed (N = 188, Group 1). To assess the effects of antiepileptic drug (AED) taper, the total AED load was calculated as the sum of the ratios of prescribed daily dose and defined daily dose of all AEDs per VEM day and was correlated with the occurrence of focal seizures without bilateral tonic-clonic seizures (FwoBTCS) and FBTCS. To validate the findings, data of patients undergoing VEM in 2004 and 2005 (Group 2, eligible N = 243, analyzed N = 203) were also investigated. RESULTS: In Group 1, 53 patients had FBTCS and 135 patients had exclusively FwoBTCS during VEM. Reduced AED load at seizure onset was the most important modifiable risk factor for FBTCS (receiver-operating characteristic [ROC]: area under the curve [AUC] = 0.78). Furthermore, the risk of FBTCS varied with the history and frequency of FBTCS prior to VEM. For instance, patients had a 50% risk of FBTCS by reducing the AED load to ~20% when no information about history of FBTCS was taken into account, to ~30% when a positive history of FBTCS was taken into account, and to ~50% when a high frequency of FBTCS prior to VEM was taken into account. These findings were largely replicated in Group 2 (59 patients with FBTCS and 144 exclusively with FwoBTCS). SIGNIFICANCE: The risk of FTBCS during VEM depends on the history and frequency of FTBCS prior to VEM and is particularly associated with the extent of AED reduction. Our data underscore the need for appropriate tapering regimens in VEM units.

Controlling decision errors with minimal costs: The sequential probability ratio t test.

For several years, the public debate in psychological science has been dominated by what is referred to as the reproducibility crisis. This crisis has, inter alia, drawn attention to the need for proper control of statistical decision errors in testing psychological hypotheses. However, conventional methods of error probability control often require fairly large samples. Sequential statistical tests provide an attractive alternative: They can be applied repeatedly during the sampling process and terminate whenever there is sufficient evidence in the data for one of the hypotheses of interest. Thus, sequential tests may substantially reduce the required sample size without compromising predefined error probabilities. Herein, we discuss the most efficient sequential design, the sequential probability ratio test (SPRT), and show how it is easily implemented for a 2-sample t test using standard statistical software. We demonstrate, by means of simulations, that the SPRT not only reliably controls error probabilities but also typically requires substantially smaller samples than standard t tests and other common sequential designs. Moreover, we investigate the robustness of the SPRT against violations of its assumptions. Finally, we illustrate the sequential t test by applying it to an empirical example and provide recommendations on how psychologists can employ it in their own research to benefit from its desirable properties. (PsycINFO Database Record (c) 2020 APA, all rights reserved).

Homogeneity of item material boosts the list length effect in recognition memory: A global matching perspective.

Kinnell and Dennis (2012) showed that the list length effect in recognition memory is only observed for homogeneous stimulus material. On the basis of the global matching model MINERVA 2 (Hintzman, 1986, 1988), we offer a theoretical explanation for this finding. According to our analysis, homogeneous material immunizes against the disruptive influence of preexperimental items, which might mask the intralist interference predicted by global matching models for familiar heterogeneous material. We tested our approach in three experiments. In Experiment 1, we found list length effects for homogeneous photographs of flowers and landscapes. In Experiment 2 and 3, we presented heterogeneous photographs of scenes (Experiment 2) and faces (Experiment 3). List length effects were only found if these photographs were homogenized by the use of image-processing filters. We further show that our explanation is also in line with the results of Dennis and Chapman (2010) who found an inverse list length effect. Overall, our results provide evidence for a global matching account of familiarity. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Assessing and correcting for regression toward the mean in deviance-induced social conformity.

Our understanding of the mechanisms underlying social conformity has recently advanced due to the employment of neuroscience methodology and novel experimental approaches. Most prominently, several studies have demonstrated the role of neural reinforcement-learning processes in conformal adjustments using a specifically designed and frequently replicated paradigm. Only very recently, the validity of the critical behavioral effect in this very paradigm was seriously questioned, as it invites the unwanted contribution of regression toward the mean. Using a straightforward control-group design, we corroborate this recent finding and demonstrate the involvement of statistical distortions. Additionally, however, we provide conclusive evidence that the paradigm nevertheless captures behavioral effects that can only be attributed to social influence. Finally, we present a mathematical approach that allows to isolate and quantify the paradigm's true conformity effect both at the group level and for each individual participant. These data as well as relevant theoretical considerations suggest that the groundbreaking findings regarding the brain mechanisms of social conformity that were obtained with this recently criticized paradigm were indeed valid. Moreover, we support earlier suggestions that distorted behavioral effects can be rectified by means of appropriate correction procedures.