Profile

Cover photo
Sanjay Srivastava
Works at University of Oregon
Attended Northwestern University
1,644 followers|21,267 views
AboutPostsPhotosVideos

Stream

Sanjay Srivastava

Shared publicly  - 
 
Are claims about being "counterintuitive" more about substance or rhetoric? New blog post...
2
Brian Knutson's profile photo
 
Pure aesthetics (assuming priors are cultural). 
Add a comment...

Sanjay Srivastava

Shared publicly  - 
1
Brian Knutson's profile photoShravan Vasishth's profile photo
2 comments
 
I know one person in linguistics who throws out data points that "don't fit the expected pattern". Hopefully this is the only one.
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
Unstructured interviews are widespread in job and school selection, even though the evidence for their effectiveness is weak to nonexistent. A new study probes why people believe so strongly in the usefulness of unstructured interviews. Here are my comments on the study and interpretations of it...
4
Sanjay Srivastava's profile photoChristopher Scollo's profile photo
2 comments
 
I agree that predicting GPA is different from predicting job performance, and if this were the only relevant study I'd be skeptical too. But in personnel selection research there have been many studies of what predicts actual on-the-job performance. There are a variety of ways to measure job performance -- some studies use relatively objective measures (such as dollar value of output), others use expert judgments (such as supervisors' evaluations), and there are other criteria too. Here's a review of studies that have used on-the-job economic output as the criterion:
http://mavweb.mnsu.edu/howard/Schmidt%20and%20Hunter%201998%20Validity%20and%20Utility%20Psychological%20Bulletin.pdf

Also, it's not so much about interviewing vs not as it is about what kind of interview. One of the big distinctions is between structured and unstructured interviewing. In a structured interview you would determine in advance what factors would make a candidate suitable for the job, design an interview that probes those topics, give every applicant the same interview so you have complete information from everybody, and evaluate everybody on all of the factors that you determined to be important. The research shows structured interviews can add useful information to the selection process (they have "incremental validity" in psychometric lingo), and they are more effective than unstructured interviews.
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
Control groups in intervention studies need to match subjects' expectations, not just activities. You'd think psychologists, of all fields, would know that.
 
New blog post: What can we learn from an intervention study?

This post starts with a pop quiz -- what is the purpose of a double blind intervention design? I then argue that almost no interventions in psychology meet the standards necessary to claim that an intervention had a causal effect on performance. The arguments are based on a new paper my colleagues and I published today in Perspectives on Psychological Science (the post has links).
1
Peter Gunn's profile photoDaniel Simons's profile photo
3 comments
 
You might be right. I guess I'm loathe to assign motives or reasons for why people have been neglecting these issues. It could be inattentional blindness of some sort (although these effects of expectations are expected, so maybe not...). It could be willful blindness. It could be that funding agencies prevent researchers from spending the money necessary for adequate control conditions (I know that has happened in some cases). If our paper forces researchers to confront these issues, that'd be great. Perhaps that can happen indirectly by getting reviewers and editors to raise these issues. If people can't publish studies without appropriate controls, they'll start using appropriate controls.
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
Option 3. Scientists produce new knowledge, and that's what they should be evaluated on. Grants are not that - they are but a means to that end. If both scientists work in a field where grants are necessary to do productive science, then the difference will show up on the next merit evaluation. On the other hand, if one person can be productive by getting grants and the other can be equally productive by being resourceful, then they are on equal footing.
Warning: This post likely will not be of interest to anyone outside of academia.  My department is in the midst of our annual faculty evaluations. Each year, our representative "advisory" committee evaluates all of the facult...
5
Joshua Wiley's profile photoSanjay Srivastava's profile photoDaniel Simons's profile photo
5 comments
 
All excellent points, +Sanjay Srivastava. I won't be solving the apples/oranges problem, unfortunately :-).I might not finish my 'further thoughts' post today, but I'll get it up soon. Lots of really interesting discussion!
Add a comment...
In their circles
244 people
Have them in circles
1,644 people

Sanjay Srivastava

Shared publicly  - 
 
Malcolm Gladwell's 10,000-hour rule has been in the news lately, so I went back and re-read what he actually said about it...
2
1
Sanjay Srivastava's profile photoSusannah Paletz's profile photoMichael Tufekci's profile photo
2 comments
 
Thank you!
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
A Very Serious Statistical Analysis comparing relative hotness and intelligence among academic disciplines.
6
1
Paul Minda's profile photo
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
First blog post in a while -- on the flawed logic of chasing large effects with small samples.
2
1
George Bowles's profile photo
Add a comment...

Sanjay Srivastava

Shared publicly  - 
 
Where are NIMH's priorities headed under RDoC? Is it significant that one of two RDoC-based FOAs issued so far is for eating disorders and the words "social," "peer," "media," "culture," "family," and "self" do not appear anywhere in it? New blog post...
1
Add a comment...
People
In their circles
244 people
Have them in circles
1,644 people
Work
Occupation
Professor
Employment
  • University of Oregon
    Associate Professor, Department of Psychology, 2004 - present
  • Stanford University
    Postdoctoral Research Scientist, 2002 - 2004
Contact Information
Work
Email
Story
Introduction
I try to figure out what makes people tick.
Education
  • Northwestern University
    B.A., Psychology, 1991 - 1995
  • University of California, Berkeley
    Ph.D., Psychology, 1996 - 2002