Not Necessarily a Smooth Ride

I teach politics part-time and my students cannot pin down my personal beliefs because I really do argue as many sides of an issue as I can. When I eventually achieve F/T status at a community college, tenure will be vital to the way I do my job. In addition to having collective control regarding issues such as curriculum delivery methods, I all too often have to present ideas that are controversial to some or even a majority of people in order that I fairly represent the broad spectrum of opinions in the public arena. If that freedom is sacrificed because I feel the need to bow at the altar of playing it safe, then I have done a disservice to my students and my discipline.

I agree with Vice Chancellor Thrift that there are significant and often unanticipated impediments relating to global student mobility–but increasing demand and numbers will mean that student mobility will continue–and probably accelerate although patterns and flows may change.  Market forces dictate it.

It seems to me that international student mobility isn’t nessesarily slowing but that the patterns of where they go may be shifting.  While the UK, for example, may see a slowing of international students due to visa restrictions, they may also see a slowing due to increased tuition, and I’d be interested to learn if there are any implications for UK nationals who may be less likely to now attend post-secondary institutions due to the rising costs as well.

As a staff member in the international student recruitment unit of my university in Canada, we definitely hear the argument that international students are taking up spaces in classes that would otherwise be reserved for Canadian students.  The only support for this argument is a statistical one, in that our entrance averages are based on a combination of the number of spaces available, the average scores achieved by all applicants to a program in the previous year, and the normal uptake of admissions offers.  The distribution of GPAs, not surprisingly, waivers very little from year to year, perhaps a percentage point or two one way or the other at most.  International student applications have the ability, certainly, to skew the averages higher if one can assume that the majority of such applicants are high achievers, but the more likely case, statistically speaking, is that the higher the volume of applications, the more normalised a distribution becomes, not necessarily skewed.  As a result, any student who meets our minimum average is guaranteed to get in, regardless of that student being Canadian or international, and as 22280998 states, international students are often more likely to study in programs that local students tend not to.  I’ve never heard of a case at our university, which has no international or domestic student quotas, where a Canadian student lost a spot because an international student got it.  The only reason any student loses a spot is that they don’t meet the minimum requirement.

In addition, our labour market will continue to require us to attract international students because as a nation with stagnant birth rates and increasing ages of others, we continue to experience labour shortages in all types of fields, from professional to labour positions, and as long as there are people who continue to want to immigrate here, I can’t forsee any change in Canada receiving international students, regardless of any protests Canadians may have.

I found this commentary interesting and i am just beginning to learn about international education.   What types of degrees are being sought by international students? Are they being accepted into program/majors in which nationals are unable to meet the requirements and this is why so many international students are taking their “space”?
Niki

As an opponent to the current acceptance levels of international students, I have some issue with your comment.  Most of us who oppose the current situation, especially in the USA, don’t mind international students paying the same as national students.  We don’t even mind if they get scholarships (although some might argue against me on that). What we are tired of is international students getting places in top US universities when they are hardly qualified.  Like I said in another post above, an Americans student needs high test scores, high class scores, work experience, volunteer experience, sports, art, music, etc.  International students, for the most part, just need to be international, have a different skin color, or a simple test score.  Most international students in the US are less qualified than American students, but the liberal argument for diversity leads to discrimination against Americans.  

How sad is this.

Reality checks are never a bad thing, even if they bruise some egos along the way.

Why is it sad?  If research is fundamentally sound then no one should have to worry about someone else reproducing that research.  Those who produce careful, rigorous research have nothing to worry about here.

All caveats aside, why is it sad to verify the conclusions of original research?  Should we merely take Newton’s gravity for granted or should we test it and try to verify that it exists? 

So much emphasis in tenure hearings tends to focus on publishing.  The gold standard is to be published in an elite “peer-reviewed” journal.  So why is it somehow bad to check that the “peer-reviewed” literature is any more accurate than something different?  There is a kind of filter on elite peer-reviewed journals that psychologists of all people should understand.

I was a student in economics.  Do you know what the biggest complaint was about getting published in that field that I heard?  The same guys publish the same stuff by getting it “reviewed” by their pals who serve on these review committees.  It is a rather small club when you start looking through the “big 5″ economics journals.

Nothing sad here.

To avoid the bias of the “file drawer effect,” journals could welcome a “negative findings” section with links to a website where readers could gather the important details of studies where “no correlation was found” and “the null hypothesis could not be disproved.” There is probably just as much value to these studies as the more spectacular ones since at least we would know what didn’t work.

Very well said.

Now try applying this logic to climate science and watch how fast you’re thrown under the bus as a flat earth bumpkin or oil industry shill.

“Like most researchers, Nosek is interested in advancing his own research
agenda rather than simply running someone else’s experiments.”

Dr. Nosek’s endeavor is not an alternative to doing research. It IS research.

Tom -

I am disappointed with the negative tone of this post.  Reproducibility (“checking your work”) is a central pillar of science.  Conducting a replication is not a threat to the integrity of the publishing scientist or of the original work.  It is an ordinary practice to improve the confidence in scientific knowledge.  The scientist ideal is not to “be right” it is to “get it right.”  

Further a failure to replicate does NOT mean that the effect is false or that the original researcher did something wrong.  It *might* mean that the effect is false, but there are many contributors to a lack of reproducibility, and all of them are important to understand.

Most important, the Reproducibility Project is a project of psychologists about psychology.  We are investigating ourselves.  The sample of studies is just that – a sample.  From that sample, we hope to learn something about the population from which we drew it.  That population includes me and my collaborators.  If the results suggest low reproducibility, then I take it to mean as much about me and my laboratory as it does about any article in the sample.  

Science is self-critical.  If there are problems, scientists aim to identify them and do something about it.  That is what I love about science, and why I am a psychologist.

I am a marketing person so perhaps I might look at this a bit differently.

1.  OK Tom good title and good introduction  you got us hooked, my guess is you will get lots of comments.  I don’t  know if you actually took Marketing 101 in college, but you are practicing it.
2.  As a marketer with a psychological bent I look at this whole issue of replication from an external validity perspective.  One of the problems we have in our field and others in the social sciences I try to read is it is difficult to get replications published.  I have done some scaling work in my career and have learned, sometimes with pain, the value of checking out the external validity of a scale based on multiple uses before I try and borrow it for my use.  While we can often find mutliple uses of a scale and the reports on how it works, we see less opportunities for replication of basic internal validity findings of specific construct relationships.  Replications only occur when we  add something to  the mix; straight replication seems to rarely be published.
3.  So I applaud Dr. Nosek for  his creative way of approaching the replication problem, really thinking outside of the box.

If this undertaking involved any vertebrates other than humans, it would probably wouldn’t be allowed by institutional animal care and use committees.  It would be deemed a waste of animal subjects.  Too bad.

Using the scientific method doesn’t make something science.  Using the scientific method is only appropriate if the thing you’re studying has aspects that can be uncovered by using the scientific method.  If the thing you’re studying isn’t consistent with the underlying presumptions of the use of the scientific method, then the scientific method is not something that should be applied.  And any results obtained thereby are suspect.

This all sounds more confusing than it is in my head…(maybe I need a psychologist..)  One way of thinking about it is that the scientific method assumes that there is no “memory” of past activities when an experiment is performed on an object.  If I do the same thing, under the same conditions to the same object, I should get the same result.  Or if I do the same thing under the same conditions to an exact duplicate of the original object, I should get the same result.  I would say that this disqualifies its use on any object with consciousness since you can never regain the same conditions on an object, nor can you have exact duplicates of two objects with consciousness.

not if it takes 20 attempts (studies) until you get a positiv and then this one gets published.

as bghansel already pointed out, the problem is not the methods, the topic, the researchers or whatever. the real problem in my view is that we still rely uniquely on gatekeepers to publish. the real bias is coming from the journals publishing only a percentage of all the studies. its not appropriate. why isnt it already common to publish first to something like arxiv.org befor you submit your paper at a journal? journals shoulnt be GATEKEEPER but CURATORS!

In my field (education), I have typically found that conducting longitudinal studies is the preferred method of gathering data and validating results for research when seeking publication. When the social sciences are involved, single studies are often simply snapshots of a particular set of characteristics for a specific group at that time. If results differ with changing group composition over 3-5 years, it provides a much broader data base from which to extract results, and derive appropriate conclusions.

 This is precisely the reason why psychology departments are self-conscious: we’re dismissed as “not a science” by people who don’t understand how psychological experiments are conducted. Most studies are replicated on new samples of respondents, for example. Personally, I applaud the replication project. If the results don’t stand up to solid replication, that’s important information. Had psychology been the first field to be selected by a team of non-psychologists, however, I would have been suspicious of motivations. Let’s also see this expanded to other fields!

A most interesting article, and also interesting comments.  I’d like to pick up some threads in the comments by tdr75, bghansel, adrianoesch, and fdhorowitz.
 
Seems the incentives are obvious.  To get tenure, one must get published.  To get published, one must have positive results — despite the persuasive case for negative results also being valuable.

Hence, there is a clear motivation to make the subjects behave in the manner desired by the one running the experiment (my paraphrase of a comment by B.F. Skinner).
 
If the status quo is unsatisfactory, I’d offer the following modest proposal for improving it.  Universally adopt the following principle:  Reproducing others’ work “counts” (in some real sense) for promotion and tenure – with, of course, confirming results counting the same as negative results.
 
 

 then they would unfortunately learn the wrong thing: alpha=0.05 means that if there is no true effect in a population (that means absolutely no effect) and given some information about variation, sampling error alone will create a significant effect with p=0.05 based on sample observations. As “exactly no effect inte population” is for most experiments a false assumption, p=0.05 does not refer to any real-world probability. For replications the situation is even worse: here you have to calculate the power of the study which you can only do if you know the true effect size. Conservative analyses suggest that many studies in psychology are heavily underpowered, though. You might well expect a higher proportion of non-replications based on power-problems alone, but you cannot calculate the probability.

Then there is the case of Jan Hendrik Schön, the prolific star physicist. Other physicists couldn’t replicate his spectacular results, to little effect, but then someone reading his high profile papers reported anomalies, which eventually lead to his downfall.
 
As usual, no fault was found with his many co-authors and peer reviewers. There were many, since he was publishing a paper every 8 days, on average, in 2001.  
http://books.google.ca/books?id=ToALIM28uScC&pg=PA27&dq=%22Jan+Hendrik+Sch%C3%B6n%22+%22eight+days%22&hl=en&ei=tFPdTrWIB6Tl0QGJ47SSDQ&sa=X&oi=book_result&ct=result&resnum=2&ved=0CDMQ6AEwAQ#v=onepage&q=%22Jan%20Hendrik%20Sch%C3%B6n%22%20%22eight%20days%22&f=false
http://books.google.ca/books?id=dP7oKntCUUUC&pg=PA144&dq=%22Jan+Hendrik+Sch%C3%B6n%22+%22eight+days%22&hl=en&ei=tFPdTrWIB6Tl0QGJ47SSDQ&sa=X&oi=book_result&ct=result&resnum=3&ved=0CDgQ6AEwAg#v=onepage&q=%22Jan%20Hendrik%20Sch%C3%B6n%22%20%22eight%20days%22&f=false
http://yclept.ucdavis.edu/course/280/Schoen.Yin.pdf

What has changed? 

Jan Hendrik Schön’s mistake was to publish very high profile results in an area of intense industrial/commercial interest which would result in experiments trying to replicate.

“Further a failure to replicate does NOT mean that the effect is false or
that the original researcher did something wrong.  It *might* mean that
the effect is false, but there are many contributors to a lack of
reproducibility, and all of them are important to understand.”

Not the least of which is the possibility that something important to the outcome was unrecognized and unmeasured in the original study — and that unmeasured factor was not common in the population of the replication study.

Unable to replicate could be a generator of new hypotheses.

So what. What if there were no field of psychology? How would humanity lose (other than an absence of an easy major for women to be sucked into so they’ll have no career future)? At a huge percentage of universities, by far the largest major is psych, despite psych grads consistently earning the second lowest. Worse yet, psych specialty fields like industrial organization and environment psych have become a dominant portion of business programs, often taught by psych PhDs out of psych texts — the core message being to bamboozle customers, employees, and suppliers into acting against their own interests. In education, psych has played a large role in the scadalously irrelevant curricula of education schools. And finally, the phoney ed leadership doctorates (chock full of low-level psych methodology) has polluted the ranks of high school prinicipals and college admistrative flunkies.

tdr75, it is joyous news that psychologists recognize that their papers are in large measure worthless.  It seems like proof that for most psychologists it doesn’t matter if what they publish is true or not, otherwise there would be no need to have a special effort to reproduce the results, because if truth mattered to psychologists and funders there would already have been multiple replication attempts.  

This activity ignores the aspect of poor statistical analysis which results in the  “plug-and-chug” approach and publication on p< 0.01, in some cases without much attention to whether the results are of any practical significance, instead of "statistical significance," which can be achieved via having a large enough sample.  

It sometimes seems you could publish that eye of newt provided statistically significant improvements in the CLA scores of undergrads, without people paying much attention to the practical significance.  It may be that the statistically significant result is only a 1% difference in means.  It might mean that 33% of students did worse, 33% stayed the same and 34% did better.  CLA itself might be a highly dubious measurement and your experimental design might have been questionable.

tdr75, I agree in general with your comments, I am just adding some additional aspects, which are all geared to showing the present deficiency of psychology compared to the history of the physical sciences (at least as taught in high school many decades ago).

Because Newton had tenure and was naturally secretive, it took Halley to cajole Newton into coming up with his Theory of Gravitation, or at least writing something down.   This theory was in fact quite controversial in some circles since it involved a mysterious “action at a distance” and failed to explain its source. 

It was largely a mathematical theory that provided a mathematical rationale behind Kepler’s results.  It is probably the Halley’s Comet return prediction that was most convincing, since it provided a  prediction that was proven true. Newton provided Physics you could use to do calculations that provided predictions that could be checked.

Henri Poincaré discovered that the Solar System could not be proved to be stable. Newton can’t even be used to solve the three-body problem except approximately (that is partly why space shots have to do “course corrections”). Henri Poincaré even discovered his proof that the Solar System was stable, was wrong and told people. He was still awarded the prize

Anomalies in the orbit of Uranus result in the discovery of Neptune and other anomalies resulted in the discovery of Pluto.  Anomalies in the orbit of Mercury failed to result in the discovery of Vulcan, but instead resulted in Einstein’s Theory of General Relativity which modified Newton’s classical theory.

I take issue with the claim that psychology does not have “aspects that can be uncovered by using the scientific method.” Your main argument against this seems to be only that human behavior is noisy and that individuals learn from repeated exposure to the same conditions. The scientific method, applied to human behavior allows us discover important patterns in how people respond to conditions. These patterns do not allow us to perfectly predict individual behavior, but they do help us understand how individuals are most likely to respond. This knowledge allows us to decide how to teach people new skills, help them make better food choices, rid them of debilitating phobias or overcome severe addictions. To dismiss the application of the scientific method in psychology because it doesn’t lead to perfect predictions at the individual level is to dismiss a large body of research findings that can and have been used to dramatically improve people’s lives.

Replication in psychology does not mean putting the exact same person in the exact same situation, because that is impossible. However, if similar people are put in a similar situation and, on average, respond the same way, then that suggest that the underlying theory is supported. Psychology is often accused of having “physics envy” but the only reason for this envy is that it would satisfy critics like willismg who fail to see value in theory that doesn’t lead to perfect prediction at the individual level.

darccity, psychology provided a vital source of “bird courses” like “Abnormal Psychology” for engineers, at least in earlier decades.
 
My guess is that in the era after reading serious novels died out and before reality television triumphed that psychology provided a painless way of trying to gain some insight into the behavior of people.   It was also probably aided by the fashion for Freud, Jung and others.  The enormous success of pop psychology, therapy and self-help probably helped.

You wrote, “bamboozle customers, employees, and suppliers into acting against their own interests. ”

Propaganda, PR, Psych Warfare, Advertising and Marketing are areas where psychology theories are tested. I’m not sure how effective they were. Goebbels and Leni Riefenstahl don’t seem to have been psych majors. Edward Bernays did come out of the Wilson Administration’s WW I propaganda machine to create “public relations.” But it is unclear to what extent the successful just use theory as a shiny gloss to sell their personal intuition.

I think those who want to dismiss the social sciences including psychology as outside of  science are invoking an incredibly scientistic approach to the scientific method. In the hard sciences the objects of our inquiry are PERCEIVED as acting as if they themselves are not engaged in sense making but that is not always (ever?) the case in the social sciences. Not only are we sometimes engaged in ignoring and controlling the sense making activities of others, sometimes we actively dismiss the sense making activities of those we study and at other times in other fields and disciplines we work to try to understand the sense made by others. In my own field (sociology) some of us are even engaged in understanding our own sense making. In any event when we work with sense making subjects and when we treat ourselves as constructors of meaning and not simply as reporters and readers of texts written by nature we can still do science but our subjects are rich, intricate, and very complex.
And those in the social sciences would seem to adopt incredibly scientistic approaches when they invoke the same kind of accounts (of literal causality and the like) as those in the hard sciences. It is the coroner who “causes” the suicide, the hospital committee that “creates” the medical error, the CDC that makes us “obese” – a very different way of thinking about “causality” than the ways that physicists and biologists think about causality.
 

pffft. This will just confirm that most psych studies are underpowered. Average beta of psychology studies is about 0.52. Thus, I expect they will replicate something less than 50% of the published works. No big deal if you understand stats. Nothing to see here that a careful meta-analysis of power in the same journals wouldn’t demonstrate (at most, the replication project will provide more reason to report effect sizes in studies, rather than only p-values).

see this youtube for why replicability is not related to p, but beta! http://chronicle.com/blogs/percolator/is-psychology-about-to-come-undone/29045

What’s to say he won’t just fake his results just so he doesn’t make psychologists look bad. That’s probably his whole reason for doing this project. He’s just gonna make all the research he checks seem real on purpose.

As a psychology researcher, I cannot express how excited I am about this. We absolutely need this to be going on. 

Hopefully, similar undertakings will occur for biomedical research, and folks will turn their eyes on the natural sciences as well – I know from experience that they share this problem with us, though fortunately it seems it is not to the same degree. 

We have a responsibility as scientists to hold ourselves and our colleagues to a far higher standard than we have.

 My experience has been that almost any field that uses the word “science” in its title doesn’t really have much to do with science…  Political Science, Computer Science, Social Science, etc….

Extending the definition of “science” so that the original intent is derided as being too scientistic (is that a word?) is tantamount to claiming that “when I use a word it means exactly what I say it does.”  (Apologies to Lewis Carroll).

In a lot of fields, 95% of what is published is not worth reading.  Face it, if it wasn’t worth attempting to replicate by a separate team, it wasn’t worth doing in the first place, let alone publishing it.

This is a facet of the way “publish or perish” has worked out.  It was already seen as a serious problem way back in 1964 in a popular science book published by Life Magazine. 

“The Scientist,” by Henry Margenau and David Bergamini:

“Antidotes for a flood 

While some scientists see automation as their last-ditch hope for coping with the mounting flood of words, others argue that the best way to control the output is to cut it down. In many universities and industries, promotions hinge, in part, on the sheer bulk of a scholar’s published papers. The pressure to ‘publish or perish’ clutters scientific literature with items of small consequence masquerading under pretentious titles. Authors of scientific papers, say the reformers, must learn to police their own productivity, and scholarly reviewers — who traditionally do not attack another man’s work unless it contains actual errors— should go further, praising genuine contributions and puncturing empty ones.”

“Plus ça change, plus c’est la même chose”
“The more things change, the more they stay the same.”

- Alphonse Karr, in 1849

And why have things not improved in the years since 1964 and instead worsened? 

We might look to: 

“Cui bono?”
“To whose profit?”

- Cicero

What’s the condition of a group of individuals every one of whom has come to expect others, outsiders, to pony up for re-examinations of conclusions…because those conclusions have no use, serve no purpose, and will otherwise never be tested or even known to exist?

The pursuit (psychology) *might* have something going for it if the experiments were assigned, similarly to the way experiments are assigned to those who engage in real science (or an art that results in either life or death): medicine.  

What if everybody’s funding depended upon someone’s discovering a link between 9/11 and Iraq, so that there’d exist evidence in favor of the war (if not in favor of a brain in every body that’s held down a chair in Congress since 9/11), which war has taken the lives of thousands of individuals…a couple of whom, if they weren’t exactly pulling for psychology, may’ve been willing to feed a few psychologists.

Not a fan of the direction the comments took. I think double checking work is always important and if someone has the time to do that then more power to them. With anything there are going to be different outcomes whether it be a cancer research study or psychology but it’s important to get the information right. So I say bring it on mass peer review!

Measurements can be accurate and not precise, precise and not accurate, both accurate and precise, or neither accurate nor precise.

How can anyone discover “true stuff” in the event trials are merely reproducible?

Most of us in the physical sciences know that science itself runs amuk when it claims to establish “absolute” truth no matter how reproducible our trials.

Determining outright fraud is one thing, but I sense aspirations to (physical) scientism by this Reproducibility Project which itself appears to labor under false premises.

“Not the least of which is the possibility that something important to the outcome was unrecognized and unmeasured in the original study — and that unmeasured factor was not common in the population of the replication study.”

Isn’t this just a longer way of saying that the original study, if it asserted there was a connection between some cause and some effect, was wrong?

I’m a sociologist, not a psychologist, but don’t psychology papers usually have the form “We did X, we observed Y, and we conclude Z”.  If somebody else does X and doesn’t not observe Y, then Z has been _disproved_ if the second study was done properly.

Yes, there might be a confounding effect.  Yes, much can be learned this way.  But the conclusion of the original study is now known to be wrong.

For evidence that in 1976 Berkeley statistics students were intellectually superior to most of the statistics TAs that were inflicted upon them:

http://chronicle.com/article/Negotiators-Reach-an-Impasse/131534/#comment-496668008

For the parallels between SAT/ACT based admissions and Tryon’s “maze-bright” rats:
http://chronicle.com/article/Harvard-by-Lottery/131322/#comment-484775851

For an illustration of how poor a predictor of college GPA that SAT can be, even with a SAT Math range of 300 to 800, for a single year, at a single 4 year university, for all Computer Science majors for 3 semesters, see the subthread starting here:
http://chronicle.com/article/Harvard-by-Lottery/131322/#comment-483721868 

SAT Verbal was worthless. SAT Math predicted 6%, while High School Math marks predicted 19% of the cumulative GPA.

For an illustration of the parallels between NCLB and the Revised Code of 1862:
http://chronicle.com/blogs/innovations/5th-avenue-percussions/30953#comment-372249648

For an example of how college seems to cause loss of civic knowledge, at least according to the ISI’s tests, please see this subthread:
http://chronicle.com/article/College-at-Risk/130893/#comment-450329553

For reasons why ed schools should not be trusted, see various comments starting here:
http://chronicle.com/article/Negotiators-Reach-an-Impasse/131534/#comment-496611616

For my critique of the misuse of medical admissions statistics to promote the Humanities:
http://chronicle.com/article/The-Liberal-Arts-as-Guideposts/130475/#comment-430281277 
http://chronicle.com/article/The-Liberal-Arts-as-Guideposts/130475/#comment-430829706

These might all be examples of why some results are not tested or replicated. People aren’t interested in the answer or don’t want the answer. It gives the impression that if something provides an undesirable answer it best that no one know about it or that it be vigorously spun to support a particular agenda.

More inconvenient info:

Trying to dispell STEM myths:

http://chronicle.com/blognetwork/castingoutnines/2011/12/07/what-makes-kids-want-to-become-engineers/#comment-383088567

As a social psychology grad student, I see nothing wrong with this. If someone is willing to replicate a study, I say kudos to them. As long as they replicate it exactly the way the original author intended the experiment to be run. The “sad” part that Dr. George Curious mentioned might be that they are targeting social psychology. Sure, that makes me a bit sad too. All research disciplines should be targeted. But alas, as tdr75 mentions, if your research is ethical and sound, why worry? 

i hope they go after the pharma journals next!

Most psychology is a load…we all know that; and most psychologists only support “new” ideas in the subject that are politically correct, because, hey, many of them are affiliated with the local college or university, and they aren’t about to bite the hand the feeds them (you know…grant money and all that).  It’s all garbage.

The nature of Psychology make “exact” replication impossible. Psychology is not an exact science.

Tell that to the many people, including many psychologists, who think or act like it is a science.

Psychology, Ed. Psychology, Social Psychology, the entire profession and its contents are utter fictional nonsense. It’s a literature course and has had horrible impacts on all spheres of American life based on air. Get rid of the Psychologists and the the Political Scientists for the same reason.
suibne

If Psychologists have not been reproducing results of studies to confirm findings then it is not a science, it’s philosophy. No other science passes of findings that cannot or have not been reproduced as fact. Except for “climate change” science. But that is also a philosophy, not a science.

Tree rings and tea leaves, the stuff of fortune tellers, not scientists.

Psychology is not now nor has it ever been a science. Really, the idea is laughable. I could write more, but what is there to say? Psychology, especially in its attempt to explain human behavior and treat aberrant behavior, is nothing more than the philosophical musings of academics and would be gurus. 

Anyone who doubts the results of his study will be free to replicate it. 

If there are a variety of valid reasons for nonreplication to occur, then why would anyone waste any time actually reading these studies and journals?   Perhaps the bar should be raised for publication so that the wheat is not lost in the chaff?  What good is new knowledge if no one knows it? 

Gee.  Next, they’ll be suggesting “reproducibility studies” in climate science…  Where will it end?

This would be great, just like world peace.  The “incentive system” in many fields includes patents, venture capital for startups, textbook authorship, etc.  Nobody gets any of that for negative results.

Interesting and great thing for the scientific method