tag:blogger.com,1999:blog-6322739827777311964.post5148450017291082016..comments2024-03-29T05:56:48.403+01:00Comments on Drang naar Samenhang: Why Social-Behavioral Primers Might Want to be More Self-criticalRolf Zwaanhttp://www.blogger.com/profile/07617143491249303266noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-6322739827777311964.post-49959906382103810932014-01-22T08:31:37.241+01:002014-01-22T08:31:37.241+01:00Thanks Brent. I agree with all three of your point...Thanks Brent. I agree with all three of your points here. I also fail to comprehend these arguments. And you're right, what's so exciting about the bingo-unscrambling task?Rolf Zwaanhttps://www.blogger.com/profile/07617143491249303266noreply@blogger.comtag:blogger.com,1999:blog-6322739827777311964.post-82266370033947641622014-01-22T08:29:33.576+01:002014-01-22T08:29:33.576+01:00I have seen several instances now (that I'm no...I have seen several instances now (that I'm not personally involved in), where a journal is unwilling or reluctant to publish an unsuccessful replication of a study that was originally published there. This is clearly something that needs to change. <br /><br />I like the goalpost quote.Rolf Zwaanhttps://www.blogger.com/profile/07617143491249303266noreply@blogger.comtag:blogger.com,1999:blog-6322739827777311964.post-31737059889084081522014-01-22T02:56:24.466+01:002014-01-22T02:56:24.466+01:00Good stuff. I thought that some of the effect si...Good stuff. I thought that some of the effect size arguments in the Dijksterhuis piece were somewhat hard to follow (see his page 73). As you note, the point of the section was to try to make a case for why effect sizes should be fairly large in priming studies versus other areas of psychology. But I was pretty much lost in that paragraph.<br /><br />1. I don't get the whole "behavioral priming experiments usually use much bolder primes than semantic priming experiments" claim. Is there some way to demonstrate the construct validity of a prime? To be frank, I think the only real evidence for the boldness of a given prime is the size of some of the published effect sizes. Thus, I worry that this kind of claim is empty – it might amount to saying that large effects sizes are large. <br /><br />2. Likewise, I did not follow the idea about how stimulus materials in behavioral priming studies are often more motivating for participants than the stimulus materials used by cognitive psychologists. I thought the point was that many behavioral primes were so impressive because they seem so subtle. What is so motivating about sentence scrambling tasks with words like Florida, old, grey, rigid, bitter? Moreover, the point in these studies is often to show that participants are unaware of the primes (again see the walking study). Is the argument then about the DVs in behavioral studies?<br /><br />3. Other passages in that piece seemed to point out that behavior is complicated and multiply determined. However, this is exactly the kind of situation that would suggest the existence of fairly modest effect sizes not large effect sizes. If so “many (social) psychological phenomena are affected by people’s mood, by atmospherics, time of day, fatigue, motivation, and even the weather” (see his page 73), then why would we ever expect large sizes with behavioral primes? Thus, I think this claim about social psychological phenomena is why large effect sizes are often so implausible.<br />Anonymoushttps://www.blogger.com/profile/07461139275622668457noreply@blogger.comtag:blogger.com,1999:blog-6322739827777311964.post-86857537754432679322014-01-21T00:21:16.540+01:002014-01-21T00:21:16.540+01:00I like to think that all quantitative science is b...I like to think that all quantitative science is basically about working out how biased this one particular die we're given is. The problem in psychology is not just that we are throwing three or four other dice at the same time (which we know about and "control for", although we don't always know how many sides they have, let alone their bias), but more the fact that the table on which we are throwing them turns out to be made of dice, a lot of which are very loosely fixed and jump into the middle of the pile when the others are thrown.<br /><br />Re Feynman's comment, I wish I was as optimistic as him. In fact maybe he's right up to the point where he says the truth will come out (albeit that the original was in JPSP and the null replication has to make do with the admirable Journal of Articles in Support of the Null Hypothesis). But after that, I don't see many examples of people saying "Gosh, I was wrong, thanks for pointing that out." A rather more common reaction seems to be "Hold on a moment which I move these goalposts" - classic pseudoscience, in other words.Nick Brownhttps://www.blogger.com/profile/18266307287741345798noreply@blogger.com