Sunday, April 7, 2019

Null Results and Some Housekeeping

This is a sort of postscript to the preceding.

If I am able to do anything even remotely right in what remains of my career and lifespan, it is to advocate for making null findings and non-replications public. In an ideal world, publishers would readily accept such findings and researchers would not self-censor. This is hardly an ideal world. I have to deal with that. However, here's the thing. I don't want to go about my work with false confidence about an effect. Not only may I be wrong, but I might unwittingly lead others down blind alleys, and that is something I am not comfortable doing.

With regard to the cognitive piece of the weapons priming effect, I am perfectly comfortable if indeed there are very few null findings when aggressive thoughts are measured and that most findings are indeed positive findings. That set of circumstances would be one that validates a couple very important set of experiments that highlighted my early career. Nothing would feel better for me, and I enjoy feeling good as much as the next person. I don't want to have false hope. I also want to communicate something else that often gets overlooked in our rush to find results that are p<.05: it is possible to get null findings or a non-replication and still do everything right. It might even be arguably probable. To the extent that I can, I want to foster a research culture that encourages researchers to come public when they aren't replicating a finding, for example. Take the individual who tried to do a subliminal priming experiment where weapons were the stimuli and aggressive thoughts served as a DV. That person was convinced he must have done something wrong. From what I recall of our conversation some two decades ago, I had the impression that this person had done his homework when it came to research design and analysis. What if we missed a golden opportunity to learn something useful, simply because we existed in a period where non-replications of that nature might inconvenience the gatekeepers in our particular field? That is more than merely a shame, and one that reflects badly not on my peer but instead on our research culture.

At the end of the day, I want to have confidence that something I thought I found is real. If there is no empirical basis for that confidence, I need to know that too. When you discover a finding that is kind of novel or quirky, go ahead and put that out there. I have no qualms doing so. That also means I have to have no qualms about being shown that my finding was a fluke. That's just the way we have to work if the science of psychology is to advance. So with regard to the weapons priming research, I really do want folks to come forward if they have non-replications that I do not know about. I will be grateful, and you will help me know how to communicate with my students and with members of my local community. If there are a bunch of unpublished successful replications, let me know as well - I will remain discrete to the extent you require of me. If something I worked on in grad school is actually right on the money, I will be grateful just to know that. Regardless, we need more openness and transparency.

One more thing. I do have a comments policy, even if it has never been explicitly stated before. I have my blog programmed to allow unmoderated comments for the first two weeks after a post is published. After that, any new comments go through moderation. I do that largely because I have no idea how frequently I can blog or how frequently I can log in to this account. Comments that are on point, whether or not they agree with a position I take are always welcome. I mainly want to protect this blog and my readers from spam. So far, I have never had to worry about trolling or abusive behavior, although I will deal with such situations appropriately if they ever occur. Basic rule of thumb: don't be a prick. Seems simple enough.

The area of research that has occupied my life is one that is filled with disagreement and controversy. I am clearly now at odds with some former coauthors as to my take on what the evidence on those matters tells us. That is an interesting space to occupy both personally and professionally. I'm okay with it. Whether those involved are okay with it is up to them, and has no bearing on how I proceed as a psychologist who wishes to communicate directly with the public.I rarely get much in the way of interaction here. More of my productive interactions seem to occur through Twitter. I am okay with that as well. This is merely a space where I can work out my thoughts in a form that is longer than 288 characters, but less formally than I would in a conference presentation or article.

No comments:

Post a Comment