Mediation analysis

Glad it’s not just me…

… mediation: a crucial issue in causal inference and a difficult issue to think about. The usual rhetorical options here are:

– Blithe acceptance of structural equation models (of the form, “we ran the analysis and found that A mediates the effects of X on Y”)

– Blanket dismissal (of the form, “estimating mediation requires uncheckable assumptions, so we won’t do it”)

– Claims of technological wizardry (of the form, “with our new method you can estimate mediation from observational data”)

For example, in our book, Jennifer and I illustrate that regression estimates of mediation make strong assumptions, and we vaguely suggest that something better might come along. We don’t provide any solutions or even much guidance.

This is from a blog positing by Andrew Gelman.  He links to a paper which purports to solve the problem, but it looks Hard.

SIS and Harrods

Some great stuff in the transcripts for “Miss X”‘s testimony:


2 Q. Was there any trace of SIS having sources or contacts or
3 others employed at Harrods?
4 A. No, there were not.
5 Q. I think you did have a number of hits.
6 A. Yes, I did, in relation to gift hampers.
7 Q. Yes. I suppose we can infer then that SIS bought
8 hampers from Harrods.
9 A. Where people had purchased Harrods gift hampers, yes.
10 Q. I do not know if that will give comfort to Mr Al Fayed
11 or not.

Update. Others here:

“Usefulness” in psychometrics

“Consider, for example, instrumentalism. This view does not invoke truth but usefulness as the primary criterion for the adequacy of scientific theories and measurements. However, we are surely not seriously considering the idea that we have to rule out rivals to the hypothesis that intelligence tests are useful. The Wechsler Adult Intelligence Scale (WAIS; Wechsler, 1955) comes in a big heavy box, which is very useful to hit people on the head with, but the hypothesis that the WAIS is valid for inflicting physical injury is certainly not the kind of hypothesis we are interested in.”

Borsboom, D., Mellenbergh, G.J., & Van Heerden (2004, p. 1065).  The concept of validity. Psychological Review, 111, 1061-1071.

Some bits from Baudrillard’s “Fragments: Cool Memories III 1990-1995”

“Either you have not to be serious and seem it, or to be serious and not seem it. Those who combine being serious with seeming serious are insignificant” (p. 5)

“Is it perhaps the case that Rossif filmed animals so well only because he secretly detested them? Everyone hides some measure of cruelty towards their object. There is no point in imputing this cruelty to unconscious motivations or some trite psychology: it is a symbolic rule. Analysis is part of the theatre of cruelty. Destruction is part of the (loving) understanding of the object.” (pps. 9-10)

“I found her so beautiful in black only because I dreamt of her dead. In fact, it was because I dreamt of her as a widow. What I was in love with in her was the allegory of my own death. But I possessed that allegory physically — which is an original form of the work of mourning.” (p. 33)

“Just as haemophiliacs are unable to staunch the flow of blood, so, semiophiliacs that we are, we are unable to staunch the flow of meaning.” (p. 77)

“The ‘therapeutic window’. What a delicious term for the interruption of medical treatment! Might you perhaps hurl yourself into the void through this therapeutic window? How about a hermeneutic window from which to hurl yourself beyond meaning. Or an existential window from which to hurl yourself out of existence and the perpetual reasons for existing.” (p. 95)

“Looking in the mirror, everyone adopts a flattering pose. In front of the cash dispenser screen, everyone takes on an air of death. Such is the terrible reflection of money on a face — or rather the abstraction of money on the absence of face. The faces are those of hostages on television, which light up only when they are released.” (p. 99)

(Page numbers from the Verso 2007 edition)

A poetic gift to me

The chains you place around my heart
are welcome ones
Just my cup of tea
For the reason that your innocence
first captivated me
in the library
my heart feeling like the
of my favorite synth song
Blood pounding through my veins
And darling, you
wear that eyeliner so well
I would hold an umbrella over you
in the rain
to keep it from running.
So please don’t go
My gothic sweet
return with me
to the library
As I’ve said,
you’re just
my cup of tea.

Rachel Lynn Brody (2005), created in three minutes from the words “Gothic”, “Chains”, “Cup of Tea”, “Umbrella”, “Reason”, “Innocence”, “Library”, and “Synth”

Different notions of “effect size”

Tired of people equating “effect size” with “standardised measure of effect size”? Here’s an antidote, thanks to Shinichi Nakagawa and Innes C. Cuthill (2007). [Effect size, confidence interval and statistical significance: a practical guide for biologists. Biol. Rev. (2007), 82, pp. 591–605.]

They review the different meanings of “effect size”:

  • “Firstly, effect size can mean a statistic which estimates the magnitude of an effect (e.g. mean difference, regression coefficient, Cohen’s d, correlation coefficient). We refer to this as an ‘effect statistic’ (it is sometimes called an effect size measurement or index).
  • “Secondly, it also means the actual values calculated from certain effect statistics (e.g. mean difference = 30 or r = 0.7; in most cases, ‘effect size’ means this, or is written as ‘effect size value’).
  • “The third meaning is a relevant interpretation of an estimated magnitude of an effect from the effect statistics. This is sometimes referred to as the biological importance of the effect, or the practical and clinical importance in social and medical sciences.”

They argue in favour of confidence intervals, as these “are not simply a tool for NHST [signifcance testing], but show a range of probable effect size estimates with a given confidence.”

They also cite Wilkinson, L & The Task Force on Statistical Inference (1999) [Statistical methods in psychology journals. American Psychologist 54, 594–604]:

“our focus on these two standardised effect statistics does not mean priority of standardised effect statistics (r or d) over unstandardised effect statistics (regression coefficient or mean difference) and other effect statistics (e.g. odds ratio, relative risk and risk difference). If the original units of measurement are meaningful, the presentation of unstandardised effect statistics is preferable over that of standardised effect statistics (Wilkinson & the Task Force on Statistical Inference, 1999).”

Good stuff, this.


“… there can hardly be any question that what ‘semantics’ conveyed and conveys to the mind of the general reader is a theory of meaning, which Tarski’s theory most emphatically was not. By calling his theory ‘semantics,’ Tarski opened the door to endless misunderstandings on this point. There has been significant damage to logic arising from such misunderstandings, from confusion of model theory or ‘semantics’ improperly so-called with meaning theory or ‘semantics’ properly so-called.”
—From Tarski’s Tort by John P. Burgess

Half of Autechre on intuition and will

“Disagreements only happen when you enter the conscious world, when you try to consider things.”

“Sometimes you have to accept that you’re a product of your environment and no matter what input you want to put into that environment, that’s just your personal taste, the culmination of all your influences, being creatively voiced one way or another, like a painter might do.”

“Some people are so tied up with the whole issue of understanding — they think you need to understand something in order to like it.”

“You don’t realise where routine and schedule become habits, and then become rules. […] It’s good sometimes to remove a lot of the conscious process.”

—Sean Booth, of Autechre, an interview, Wire (2008, March)