[Rasch] current issues of research in psychometrics?

Andrew Kyngdon akyngdon at lexile.com
Thu May 3 11:30:21 EST 2012


Dear Anthony,



You say:



I looked at some journals but couldn’t ‘detect’ a specific line of research
which one can say that psychometricians are following now.



I doubt that there has ever existed any coherent line of research in
psychometrics, if by “line of research” it is meant a program of sustained
research into the psychology of individual differences in test performance
which has resulted in a steady accumulation of knowledge of such
differences.



Have a look at the website for “Psychometrika”, the flagship journal of
psychometrics. In 2010, the most viewed paper online was Cronbach’s
original 1951 paper on “alpha”. In 2006 it was the same. When I looked on
Tuesday (1st May), it was the most viewed paper in the previous 90 days.
Rather amusingly, the second most viewed paper was Sijtsma’s (2009) paper
titled “On the use, the misuse and very limited usefulness of Cronbach’s
alpha”.



Has there been nothing truly new and groundbreaking in psychometrics in
over 60 years? Is Cronbach’s alpha the single most important and
consequential idea that psychometricians have devised in all this time? It
would appear so.



Contrast this with the study of decision making under conditions of risk
and uncertainty of the past 60 years. The most cited article of all time
for “Econometrica” is Kahneman & Tversky’s (1979) paper on “prospect
theory”, which founded the field of behavioural economics. The most viewed
paper online for “Journal of Risk and Uncertainty” is Tversky & Kahneman’s
(1992) paper on “cumulative prospect theory”, which is prospect theory with
the incorporation of the rank dependent functional proposed by the
Australian economist John Quiggin (1982).



Sixty years ago we thought that people were all strictly rational utility
maximisers (i.e., all choice behaviour conformed to the von Neumann &
Morgenstern (1944) axioms) who are always risk averse and aware of how
their choices affect their total asset position. Today we know why:



·        people prefer sure consequences (gifts) over gambles which have a
high probability of winning a larger amount of money than the gift (Allais,
1953);

·        people do not tend to consider their total asset position when
making a risky choice (such as when purchasing a lottery ticket);

·        people counterintuively seek risk when faced with certain losses
(Tversky & Kahneman, 1981);

·        people’s choice behaviour is highly sensitive to the way a choice
is framed and presented to the decision maker (Tversky & Kahneman, 1981);

·        people buy lottery tickets, play poker machines AND purchase
insurance; and why strictly rational choice fails to explain this (Tversky
& Kahneman, 1992);

·        loss aversion causes the equity premium paradox (Bernatzi &
Thaler, 1995);

·        loss aversion causes the status quo bias (Williamson & Zeckhauser,
1988);

·        there is a lower rate of tax evasion than is predicted by strictly
rational utility maximisation (Bernasconi, 1998);

·        stock market levels move around too much (the “volatility puzzle”)
and the predictability of price to earnings ratios (Barberis, Huang &
Santos, 2001);

·        choices between insurance policies is not strictly rational
(Johnson, Hershy, Meszaros & Kunreuther, 1993); and why people purchase
more insurance than what they actually need (Segal & Spivak, 1990);

·        people are highly averse to “probabilistic insurance” (Wakker,
Thaler & Tversky, 1997);

·        the behaviour of options traders is not strictly rational when the
probabilities of returns are not fully known (Fox, Rogers & Tversky, 1996);

·        some investors choose portfolios that lie underneath the Sharpe –
Linter Capital Market Line (i.e., choose portfolios of stocks
stochastically dominated by others) (Levy, 2008);

·        investors tend to hold onto their losing stocks to a greater
extent than their winning stocks (Shefrin & Statman, 1985);

·        people’s willingness to bet on an uncertain event depends not only
upon the uncertainty, but upon the source of the uncertainty (the Ellsberg
(1961) Paradox);

·        political leaders take risks to maintain their international
reputations and domestic support (Levy, 1997);

·        political leaders, when they have suffered losses (e.g., domestic
support, territory) take excessive risks to recover such losses rather than
wear them (Levy, 1997); and even why

·        Jimmy Carter chose to use military force in the Iranian Hostage
Crisis of April, 1980 (McDermott, 1994).



A plausible account for all of the above is given by either prospect theory
or cumulative prospect theory. Can we point to anything like this situation
in psychometrics? Do we really know anything substantial about individual
differences in mathematics ability, for example, than we didn’t know 60
years ago?



I doubt it.



Andrew



*From:* rasch-bounces at acer.edu.au [mailto:rasch-bounces at acer.edu.au] *On
Behalf Of *Anthony James
*Sent:* Wednesday, 2 May 2012 3:45 PM
*To:* Rasch at acer.edu.au
*Subject:* [Rasch] current issues of research in psychometrics?



Dear all,

A colleague of mine asked me “What are the current issues of research in
psychometrics these days?”

I looked at some journals but couldn’t ‘detect’ a specific line of research
which one can say that psychometricians are following now.

Does really such a line(or a number of lines) exist?

Cheers

Anthony
------------------------------
This email and any attachment (“the Email”) is for the exclusive use of the
addressee. The Email may contain information that is legally privileged,
confidential or protected by copyright. Privilege or confidentiality is not
waived or lost, nor is any licence granted in respect of copyright, by
virtue of delivery of the Email to you. If you have received the Email in
error, kindly notify the sender by return email and delete the original and
any copies. To the full extent permitted by law ACER does not accept
liability for any loss or damage caused by: you failing to check the Email;
transmission errors; or your reliance on or use of any information
contained in this email. For further information refer to
http://www.acer.edu.au/about/privacy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20120503/06b4a1a1/attachment.html 


More information about the Rasch mailing list