
OK, one more thread about the AAUP and strategy. I want to give a real critique, one that's supportive of the org, concrete, and not hyperbolic. It's grounded in the idea that "fighting" and "attacking" are not always the same thing.

Assistant professor at UTK | PhD in Human Dev. & Family Science | Stress and resilience | Family policy | #Rstats | views my own

OK, one more thread about the AAUP and strategy. I want to give a real critique, one that's supportive of the org, concrete, and not hyperbolic. It's grounded in the idea that "fighting" and "attacking" are not always the same thing.

Overall, the CSI-4 did well at distinguishing people with low-to-average relationship satisfaction. But, it suggested that lower-income individuals' satisfaction was lower than it really was. Check out the paper and let us know your thoughts! (8/8) doi.org/10/m69t

@SNiehuis @AlanReifman @EmmaGrossmann et al. found that this item performed well on its own. Researchers may consider sticking to this item when evaluating satisfaction between groups. You can also use some more complex modeling procedures, outlined in the discussion. (7/8)

So what? First, researchers should always test for measurement invariance. We can't assume measures work the same for everyone. Item 4 - "In general, how satisfied are you with your relationship?" - performed similarly for everyone. (6/8)

This is a meaningful difference, given the CSI-4's distress cutoff score of 13.5. This measure is also used to test the effectiveness of programs and interventions. Fortunately, the difference between lower-income and middle-income individuals was small. (5/8)

Using item response theory and @NCHATstudy data, we found that individuals with lower incomes (<200% FPL) were systematically disadvantaged on the CSI-4. On average, at the same true levels of satisfaction, they scored 1.38 units lower than higher-income individuals. (4/8)

Research shows these approaches vary based on one's income. So we thought the CSI, which treats relationships individualistically, might perform differently across the income spectrum. (3/8)

The CSI is a great measure. But to this point, researchers have assumed it works the same for everyone. People vary in their orientations within relationships. For instance, some are individualistic, others more interdependent. (2/8)

New paper alert! Do you study relationship satisfaction? Ever use the Couples Satisfaction Index? Check out this new paper from yours truly and @JeremyBKanter on how the CSI performs across income groups 👇 doi.org/10/m69t (1/8)

Helpful new pub in Social Policy Reports on using evidence to shape policy: doi.org/10.1002/sop2.33

in the Administration for Children and Families to learn more about this exciting opportunity to apply for a grant of up to $25,000 per year for 1-2 years! grants.gov/search-results-de…

Are you a PhD candidate looking to support your research on issues facing families with low incomes in the United States using a behavioral science lens? Check out the new funding opportunity announcement from the Office of Planning, Research, and Evaluation...

What's the biggest lesson you learned during your PhD studies? Wrong answers only. (Especially insightful or funny answers may be included in a commencement speech!) #academicchatter #phdlife #academictwitter #phdchat

P.S. Probably the best part of this dissertation process was my wife's reaction when I said "heteroskedasticity" while practicing my defense.

Thanks also to @JeremyBKanter @BrendaLohman @DrTylerJamison @cmproulx2002 @bogolsky and many other Twitterless colleagues for the mentorship and support that's shaped who I am as a scholar!

Really, this is bad news, because now I can't assume that any email addressed to "Dr. Ogan" is junk mail, meaning I'll have to read more emails. But these are the struggles academics are called to overcome.