Is empathy important for forgiveness?

The authors have clearly addressed all my comments. I am happy with the study in its current form.

For more of a suggestion that applies to any exploratory analyses, I appreciate its an area of debate in replication work whether you should follow the target study as close as possible, “warts and all”. If you change the approach, it opens an argument on whether it’s a direct replication or not. For my comments on correcting for pairwise comparisons and assessing the impact of outliers, I’m happy for them to be added as additional exploratory analyses, but I would encourage exploring their impact to comment on whether the results were robust to these kind of analysis choices.

In the interests of transparency, I sign my reviews.

Dr James Bartlett, University of Glasgow, UK.

*Comments to the Author:

Many thanks for allowing me to read the Stage 1 Registered Report entitled “The impact of Empathy on Forgiveness: Replication and extensions of McCullough et al. (1997)'s Study 1”. The authors tried to replicate the classic MuCullough et al. (1997)'s Study and to extend the study by manipulating empathy. I approached it with great personal interest and greatly appreciated the efforts made for reproducibility and replicability in the psychological sciences. Overall, the paper is well written and well structured, but there are still some concerns for this paper.

While I strongly agree with the importance of McCullough's paper and the significant foundation it provides for the study and development of forgiveness. I doubt the contribution of replicating the study of McCullough et al. (1997). 

First, correlations or causal relationships between apology, empathy, and forgiveness have been tested by forgiveness researchers individually or together, and even meta-analyzed (e.g., Carlisle et al., 2012; Fehr et al., 2010; Konstam et al., 2001; Paleari et al., 2005). These efforts have to some extent replicated the work of McCullough et al. (1997), making a simply close replication does not seem necessary. I would rather suggest that the authors extend alternative meaningful models while partially replicating the study.

Second, as the field of forgiveness continues to evolve (based on McCullough and colleagues), the classification has become clearer, i.e., trait forgiveness (or so-called forgivingness; Brown, 2003) and state forgiveness (as in your case), and the term “forgiving” is mentioned less and less. I do understand that you use “forgiving” and “forgiveness” interchangeably through the manuscript, but this is not rigorous to the scientific development process. In addition, operationalizations of forgiveness measures have advanced, such as the Transgression-Related Interpersonal Motivation (TRIM), and the Enright Forgiveness Inventory (EFI) has been validated to have good internal consistency (Card, 2018). Indeed, the original source of the TRIM was Susan Wade Brown’s doctoral dissertation at Fuller Seminary (Wade, 1990). It may not be appropriate to replicate the study simply using measures from two decades ago.

Third, as you wrote in the manuscript on Page 11, McCullough et al. (1997) conceptualized interpersonal forgiving as “the set motivational changes whereby one becomes (a) less motivated to retaliate against an offending relationship partner, (b) more motivated to maintain estrangement from the offender, and (c) more motivated towards conciliation and goodwill for the offender, despite the offender’s hurtful actions”, such that the increased conciliation motivation and reduced avoidance and revenge motivation is defined in the nature of forgiveness. So there is little meaning to examining the associations between forgiveness and these motivations (hypothesis 2). Instead, it may be more interesting to investigate real behaviors/actions.

I am also concerned about testing the mediation model with a manipulated mediator (i.e., empathy). Given that mediation analysis is inherently causal, that is IV predicts Mediator and DV, M predicts DV. The original model of McCullough et al. (1997) was based on a correlation design and their limitations in causal inference are unquestionable. However, I don’t think a manipulated Mediator helps to establish the causality of the impact of perceived apology on empathy, let alone the mediation model in your Figure 10.

Other issues: 

The authors manipulated empathy with only one sentence “you were highly/not empathetic toward the person who had hurt you”. I would like to know if this manipulation is valid in previous studies, if so, please cite their work, if not, a pilot study to justify this manipulation in necessary.

The second line on Page 15 is unfinished.

Thank you for inviting me to review this pre-registration (replication study on the link between empathy and forgiveness in McCullough et al. (1997)'s Study 1.)

Overall, the process is explained thoroughly.

·         I agree with the authors that to date, there are currently no published independent direct replications of this article. McCullough et al. (1998), thus, i believe that extending their model by adding other variables such as commitment, the impact of the offense, and rumination, into predicting forgiveness would provide a decent contribution to the literature. 

·         The authors provide a sufficient justification for replicating the study, the research design is logical and addresses the coherence and credibility of the hypotheses

·         The introduction provides a good rationale and justification for the study, 

·         The sample size in the current study aimed to recruit 800, which is a well-powered sample size to predict this effect. 

·         The study procedures and analyses, and the experimental conditions are appropriate the scales also are appropriate. 

Best of luck with your research.