Childhood adversity and lifespan

 

Here is a study which deserves attention, based on an excellent sample which is dear to me and to many social scientists: the 1958 National Child Development Study (NCDS) which included all live births during 1 week in 1958 (n = 18,555) in Great Britain. If any dataset deserves a monument, this one does. For my generation of psychologists it provided a steady flow of publications on child development, so it is natural to regard it as one of the family, a familiar companion. Their 1972 “From Birth to Seven” had a great impact, serving as a benchmark for the times. Call it the thinking person’s “Seven Up” (a film which tracked a few children).

Now a new team have looked at this old warhorse again, and compared each child’s  documented early childhood “adverse experiences” with their lifespan, finding that real adversity has shortened lifespans. More specifically, a total of 3.2% of all the sample were dead before 50 and for the boys, where the adversity effect was strongest, the low rate of death doubled for those who had two or more adverse events. (The great advantage of a contemporaneous record of these experiences is that we do not have to rely on faulty and possibly biased recall in adulthood). The finding makes sense, from a trauma point of view. The child has been subjected to a dreadful event, the trauma has embedded itself during a sensitive period in development, and as a consequence children’s resources have been diminished,  and it all comes out in the end, by causing a somewhat premature end. An unpleasant finding, but a salutary one, since it shows how vulnerable children are to external events.

The effect is only slightly diminished by controlling for the effects of social class and education. In summary: an impeccable sample, distinguished authors, appropriate statistics, and a clear exposition of the effects in copious tables. It is natural that UCL should choose to mention it in their weekly newsletter, which is where I found out about it. As someone who worked for a long time in trauma, the paper is particularly welcome. Case proven.

http://link.springer.com/article/10.1007%2Fs10654-013-9832-9/fulltext.html

Before turning away, have a look at the list of adverse events.

1. Child in care: child has ever been in public/voluntary care services or foster care at age 7, 11 or 16.

2. Physical neglect: child appears undernourished/dirty aged 7 or 11, information collected from the response from child’s teacher to the Bristol Social Adjustment Guide.

Household dysfunction, as described by Felitti et al. [16], is a dimension of adversity consisting of four categories each contributing to the score:

3. Offenders: The child lived in a household where a family member (person living in the same household as the child) was in prison or on probation (age 11 years) or is in contact with probation service at 7 or 11 years; the child has ever been to prison or been on probation at 16 years.

4. Parental separation: The child has been separated from their father or mother due to death, divorce, or separation at 7, 11 or 16 years.

5. Mental illness: Household has contact with mental health services at 7 or 11 years; Family member has mental illness at 7 and 11 or 16 years.

6. Alcohol abuse: Family member has alcohol abuse problem at 7 years.

About 5% to 7% of the total sample had experienced these events, which put them in what we might call the troubled minority.

Now comes the question: which of these events are truly external? To prompt you, in case you find the question hard: what sort of parents end up with their children in care; physically neglect them; get put in prison; have mental illness; abuse alcohol; and cannot stay together when their children are young and need them? Might they conceivably be parents who are not passing on the very best genes to their offspring? Far from being external, I see these events as arising from bad parenting, and one very vivid possibility is that the offspring have been subjected to a double burden: bad experiences, and genes which make them vulnerable. Even if the authors find it hard to countenance the notion of the heritability of mental illness (see previous post below on “Bad Blood”), one expects some comment.

So, how do the authors deal with this possibility? They do not mention it.

Here is what they say about “prior confounders”: “To examine the relationship between ACE and premature death, prior confounding variables potentially associated with both ACE and mortality were adjusted for in the multivariate models. Among the variables available at baseline, collected from the cohort member’s mothers via a questionnaire at birth, we identified those most likely to be social or biological confounding factors. Household and parental characteristics were included: mother’s age at birth, overcrowding (people-per-room), mother’s partner’s social class (manual/non-manual), and if this was unavailable the mother’s father’s social class was used, mother’s education level (left school before/after minimum leaving age), and maternal smoking during pregnancy (no smoking/sometimes/often/heavy).”

As you will see, there are many things about the parents which they do not know, presumably because they were not asked in the original survey. Social class and education are certainly a help, but they are rather broad categories, and hardly a proxy for biological markers.

What else could these authors have done? Well, they could have looked at intelligence measures, since lower intelligence is associated with shorter lifespans. The children’s intelligence was measured at 11. It does not cover all the possible biological markers, but is associated with it. Finding no association at all would strengthen the trauma explanation.

Gale et al. included intelligence in a study on this sample’s smoking rates, and found “women with lower IQ in childhood were more likely as adults to smoke during pregnancy and to be a smoker currently. Structural equation modelling showed that the effects of childhood IQ on smoking behaviour were indirect, as they were statistically mediated by educational attainment and age at first birth. There was some effect of educational attainment and age at first birth on smoking behaviour over and above the effect of intelligence.”

http://www.ncbi.nlm.nih.gov/pubmed/18812363

If the data are there, why not use them? The authors of the current paper did take measures of education, social class and health at age 23 and they say “in our model, these adult variables at the age of 23 are a proxy for behavioural patterns in early adulthood and controlling for them serves as a first step to understanding possible mechanisms.” Controlling for those did reduce the hazard ratios somewhat.

What could they have done? I assume that in this study the age at death of the parents is known. If the pattern of premature death is as strong in the parents as in their children (for the sample as a whole, not just the adverse event group) that would be germane, and would weaken the trauma-leads-to-death argument.

In summary, I think that a few more checks would potentially strengthen this study. The effect they have found seems noteworthy, but they could at least have mentioned a genetic explanation. In terms of exploring alternative hypotheses,there are so much data in this goldmine of child development that it would be a pity not to complete the picture.

There is one other thing the authors could have done. They could have changed their title to: “Adverse parenting and premature mortality”

Good sample, though.

Is blood a social construct? A Challenge

 

Last year, like any good citizen past middle age conforming to public health requirements, I had a blood test, which gave me the welcome news that, having peered into this very personal essential fluid, it was judged unlikely that I was developing bowel cancer or prostate cancer. Of course, I know that no test has perfect sensitivity, but it was reassuring news.

Included in the long list of results was the following:

GFR….. if patient of black African descent eGFR must be corrected (multiply by 1.2)

This caused me no particular surprise. I had been seeing these correction factors in patients’ notes for some years. Epidermal growth factor is not something I deal with as a psychologist because, yes, I don’t like the sight of blood. Here are the NICE guidelines http://guidance.nice.org.uk/DT/11 if you are interested in it.

However, I recollected this fact a few days ago when I came across the following tweet from a UK Professor: “Race= biologically meaningless term”. I queried this, and a long exchange of tweets ensued, and is continuing.

It leaves me with a paradox: both I and the NHS accept the description “black African descent”. We know that it means people most of whom, or virtually all of whom have come from the African continent for generations, and are black Africans. It cuts out the Afrikaans, the Indians, and other racial groups that migrated to Africa in recent generations. However, in so far as I understand what is being denied by the professor’s phrase “Race= biologically meaningless term” it seems to cut out the use of this correction factor, on the grounds that drawing a distinction Africans and non-Africans is “biologically meaningless”.

So, here is my challenge: a good bottle of fine French wine to anyone who can show me and the NHS, using good quality genomic data, that it is wrong to draw any distinction between Africans and non-Africans in analysing blood.

Bad blood

Here is an interesting study, which simultaneously seems wrong and right. Wrong, because the stated risks of mental illness in offspring seem so high, and right because it soon becomes obvious that many family studies have highlighted the wrong thing: they have assumed that if a parent has a particular, individual severe mental disorder, then the main thing to look for in offspring is the same mental disorder. This limits the calculation of risk too narrowly. The better question is the one sometimes asked by patients: I have mental problems, so what is the chance that my kids will have mental problems?

Risk of Mental Illness in Offspring of Parents With Schizophrenia, Bipolar Disorder, and Major Depressive Disorder: A Meta-Analysis of Family High-Risk Studies  Daniel Rasic, Tomas Hajek, Martin Alda and Rudolf Uher

http://schizophreniabulletin.oxfordjournals.org.libproxy.ucl.ac.uk/content/early/2013/08/18/schbul.sbt114.full

Rasic et al. give their answer thus:  1 in 2 of your children will have mental problems.

How have they got to so high a figure, when the population estimates for individual psychiatric disorders are so much lower? The authors have taken the broad category of severe mental illness (schizophrenia, bipolar disorder, and major depressive disorder) and then looked for those disorders and any other mental disorders in the offspring of parents with those disorders, compared with severe and all other mental disorders in controls, the offspring of parents without those severe mental illnesses. They have looked at 33 family high-risk studies which have investigated psychopathology in individuals who had a biological relative, most commonly a parent, with SMI. These studies have provided a wealth of information about risk of mental disorders in offspring. However, the rates of mental disorders in offspring vary across studies, and a robust overall estimate is not available. They try to correct that omission. Here are the results from their Table 2 for offspring for all ages, taken from the meta-analysis described in their paper.

Parent diagnosis          Offspring disorder (relative risk)

SMI                               SMI    2.52     Any mental disorder   1.60   

Schizophrenia              SMI     2.21     Any mental disorder   1.45

Bipolar                          SMI    2.42     Any mental disorder   1.66

Depression                   SMI     2.45     Any mental disorder   1.64

 

Relative risk is not a particularly friendly statistic. It can mislead, and often requires looking up its definition. Absolute risk (which they give for all disorders in Table 1)requires less explanation  The chance of getting an SMI in the general population is 12% and the chance of getting an SMI if your parent has one is 32%, which is 2.5 times higher, thus providing the relative risk estimate of 2.52. As always, there is the problem of definition as to what constitutes a psychiatric disorder, but having a parent with a major psychiatric disorder makes it two and a half times more likely that the children will get the disorder. That is a very significant increase.

To look at the total effect for the children, one has to multiply these relative risks by the frequency of each of the parent diagnoses. Since depression is more common than schizophrenia, depression adds more to the eventual risk for offspring of getting any mental disorder.

Psychologists rarely miss a chance to criticize psychiatric disorders, but here is a result which should give pause to everyone in the psycho-professions: the broad grouping of SMI seems to be more useful in social terms than the more precise individual diagnoses (it indicates that the person is significantly at risk, needs help, and will have times, often long times, when they cannot contribute much to others) and the broad category appears to have higher predictive value for children’s outcomes.

Of course, a disordered parent passes on genes and also a disordered or interrupted parenting to their child, but the overall consequence is what matters.  This study shows how a broader approach can give a better estimate of overall risks, and helps bridge the gap between family high-risk studies and population registry studies. In a nutshell, given bad blood, something of that bad blood is very likely to pass on to your children, such that one third of them may develop a mental disorder by early adulthood. Sobering, and closer to the foul mutterings of the man and woman in the street, who always regard mental problems with some alarm and prudent avoidance.

Poverty and Class, and a mostly missing variable

 

Courtesy of the British Psychological Society, I was alerted to the recent publication of the Journal of Educational Psychology “Virtual Issue: Research from Educational and Developmental Psychology on Poverty and Class“ Edited by Harriet Tenenbaum.

The editor writes in her introduction: “The first half of the selected articles in this virtual issue focus on relations between income and education and the second set on children’s views of economic inequality. The first set of selected articles address why children from low-income families may do less well in education, but also address interventions designed to improve the education of children from low-income families.” “The studies reported in these articles were conducted in many different countries conforming that the influence of poverty on education is important in countries as egalitarian as Sweden.” “In general, the first set of these articles indicate that income serves as a risk factor for educational attainment (Cassidy and Lynn, 1991) partially because of cultural capital or class reproduction (Myrberg & Rosen, 2009). Other factors, in addition to income, influence school dropout (Fan & Wolters, 2012) and achievement (Cassidy & Lynn, 1991).”

There are 10 articles in all. The first mentions intelligence by name, and shows that it has an influence. It also measures personality. One other (Dockrell, Stuart and King) includes some British Ability Scale subtests, and another (Sylva et al.) includes the British Picture Vocabulary Scale as a measure of literacy and general ability.

So, at very best, out of 10 key papers, chosen to be of above average quality and impact on the theme of poverty and class, 3 mention ability, one of those three explicitly. That first paper provides a good model as to what every educational paper should include: a measure of intelligence (here the Differential Aptitude Test); a measure of personality (here the Junior Eysenck Personality Questionnaire); measures of social class, parental education and occupation, family size, type of school attended, an index of possessions in the home, a home crowding index, a measure of parental encouragement of education; and then the target variable, a 20 item measure of achievement motivation. As you can see, all putative causes get a chance to contribute. This paper by Tony Cassidy and Richard Lynn uses path analysis (in 1991) and concludes:

“This study replicates the findings by Lynn et al. in that school-type, IQ, and home
background are important predictors of educational attainment. School-type is the single most important predictor, accounting for 19.6 per cent of the variance. IQ accounts for 13.1 per cent and the combined home background variables of crowding and parental encouragement account for 18.1 per cent of the variance. The other important direct predictors of educational attainment are the achievement motivation dimensions of acquisitiveness, dominance and work ethic. Among them these variables account for 36.2 per cent of the variance. The total variance accounted for by the variables in this study is 87 per cent.”

However, more important than the results of a particular study done 22 years ago, is that so few of the other, usually far more recent, paper show the same even handed range of variables. Since intelligence precedes schooling, and schooling precedes getting a job, and a job precedes accumulating wealth, one might have expected at least basic measures of intelligence to be found in all these educational papers. Its absence allows the authors to draw conclusions about class, poverty, and educability without intelligence having a chance to be tested as a contributing factor. Most seriously, its absence often leads to thoroughly misleading conclusions. For example, Fan and Wolters measure school dropout without taking any measures of intelligence, a very significant omission. Dockrell, Stuart and King have obtained some measures of intellectual ability, but do not display the relationship between intelligence and learning in their intervention study, which would have been instructive.

The papers in the virtual edition are very probably a fair selection of educational psychology publications. Most of the sample sizes are small,  particularly in relation to the effects being studied, and few are population based. My main gripe is that I do not think that educational psychology papers give a fair place to intelligence assessments (nor to personality measures). It is a pity, and significantly reduces our ability to understand children’s educational progress.

A new textbook of psychology: contributions invited

 

Having had my tea, I decided it was time to write a new textbook of psychology. I decided that, to make the task more manageable, I would cut out: all historical introductions; all hypotheses however engaging which were unsupported by relevant confirmatory findings; all individual studies unless they illustrated a general point which had been replicated many times; most individual studies unless they were on substantial and highly representative samples; and all studies where either the hypotheses or the methods or the analysis of results were not clearly specified, and had not been replicated at least three times.

This is what I have got in my draft outline so far:

Chapter One: Psychological Theories which are non-trivial and well supported by results.

 

 

 

Any suggestions as to what I should include?

Breast feeding, intelligence, and confounded researchers

 

In a decade or two, people may look back on us as having lived in the dark ages, blind to the obvious, and bound up with delusions, angels and devils. Our behavioural science research designs may look hopelessly simple, and wide open to confounding effects. Full genomic analyses may go a long way to resolving some of those confounders, but good research designs will always be required.

Here is a simple question which dates back to 1929: does breastfeeding a child, as opposed to giving them formula milk, boost their intelligence? Set aside for a moment the potential benefits of bonding with the child, protecting them from illnesses, and thoroughly irritating passers-by who hate the sight of humanity, and just concentrate on that question.

Walfisch, Sermer, Cressman and Koren have done just that in a BMJ paper: “Breast milk and cognitive development—the role of confounders: a systematic review” Their abstract immediately hits the nail on the head: “The association between breastfeeding and child cognitive development is conflicted by studies reporting positive and null effects. This relationship may be confounded by factors associated with breastfeeding, specifically maternal socioeconomic class and IQ.”

They do a systematic review of the literature,  and found 84 studies met their inclusion criteria (34 rated as high quality, 26 moderate and 24 low quality).

They explain: “Well-established confounders in breastfeeding research include demographic and IQ differences between mothers who breastfeed and those who choose not to. Parents who score high on a range of cognitive abilities have children with above average IQ scores. In parallel, advantage in mother's IQ more than doubles the odds of breastfeeding. Thus, some of the published data demonstrates the disappearance of the breastfeeding effect on child's cognition after correction for maternal IQ.”

“Given that more tight control of confounders resulted in greater likelihood of disappearance of breastfeeding effect, it can be argued that the remaining positive effect reflects residual uncontrolled bias, as shown by Der et al in their large study. In that study, before adjustment, breastfeeding was associated with an increase of around 4 points in mental ability. Post hoc analysis revealed that adjustment for maternal intelligence accounted for most of this effect—where full adjustment for a range of relevant confounders yielded a small (0.52) and non-significant effect size (95% CI −0.19 to 1.23).”

They conclude: “Much of the reported effect of breastfeeding on child neurodevelopment is due to confounding. It is unlikely that additional work will change the current synthesis. Future studies should attempt to rigorously control for all important confounders. Alternatively, study designs using sibling cohorts discordant for breastfeeding may yield more robust conclusions.”

My conclusion: Breast feeding is probably a good thing, but don’t do it with the sole purpose of boosting the intelligence of your child. In all probability, the only way to boost the IQ of your child is to make a careful choice of mate 9 months before. So there’s a headline: “Mate selection more important for your child than breast feeding”. Worth a tweet at least.

 

BMJ Open 2013;3:e003259 doi:10.1136/bmjopen-2013-003259

The truth about Princess Diana: new bullet found

 

Piaget found that if you told young children a story about a boy who had accidentally broken seven cups, they regarded that boy as far naughtier than a boy who had accidentally broken only one cup. From a cognitive point of view one can explain this by saying that the children in this experiment do not yet fully comprehend the notion of an accident. If something is truly accidental, then the resultant damage was not intended, and the concept of naughtiness is inappropriate.

Of course, it is hard to shake off the feeling that a child who breaks seven cups by mistake is paying less attention than a child who breaks one by mistake and is therefore culpable, because there is a tendency to think that things must be proportional, simply because they often are.

Princess Diana was one of the most photographed people on the planet. She was avidly followed by millions. Her face on a magazine always boosted sales. Her story was a drama with which many could identify. When she died in a car crash there was an understandable wish to find out exactly what had happened, and an underlying assumption that the causes of her death would be proportional to her fame. A woman of her renown could have been tracked, followed, stalked and then dispatched by persons unknown, for reasons unknown, the complexity of all these arrangements (someone waiting by a Parisian tunnel with a laser or a sniper’s rifle) being almost a homage to her, albeit of a malevolent kind. The notion that a woman of stratospheric status could be brought down by an inebriated chauffeur is hard to stomach, and frankly somewhat insulting to her. Rather than the Just World Hypothesis, this is the Unjust World Hypothesis: if something manifestly unfair and shocking takes place, forces of unfairness and shock must have been assembled with commensurate malevolence so as to break the protective carapace of Fame. It would be good, in the twisted sense of that word, to find a bullet somewhere.

Why has there never been a conspiracy about J.D.Tippit? (I wrote his name from memory, and when I checked I found I had got it slightly wrong, which strengthens my following argument). He was the Police Officer who was shot by Harvey Lee Oswald 45 minutes after someone shot J.F.Kennedy. Both Tippit and Kennedy had served their country. Both saw active service, Tippit being in action on the Rhine and getting decorated for it, Kennedy having served in the US Navy. Tippit had been a good cop for 11 years, Kennedy an inspirational President of the the United States for 3 years. The fact that Tippit was shot by a disturbed loner makes sense. In the popular mind, it matches. Cops sometimes get shot by no-good guys with guns, and Tippit was an unknown cop. The same disturbed loner being able to kill a nation’s handsome president makes far less sense. Once again, it is pretty insulting. Surely more assassins are required to bring down a very high status person? At least one other armed man must have been somewhere, on a grassy knoll of popular indignation.

So, every anniversary a new bullet is found, some scrap of doubt which is twisted into the projectile of a disturbing fact by some people’s longing for proportionality.

Gigerenzer and psychological theorizing

 

“Several years ago, I spent a day and a night in a library reading through issues of the
Journal of Experimental Psychology from the 1920s and 1930s. This was professionally a most depressing experience. Not because these articles were methodologically mediocre. On the contrary, many of them make today’s research pale in comparison to their diversity of methods and statistics, their detailed reporting of single-case data rather than mere averages, and their careful selection of trained subjects. And many topics—such as the influence of the gender of the experimenter
on the performance of the participants—were of interest then as now. What depressed me was that almost all of this work is forgotten; it does not seem to have left a trace in the collective memory of our profession. It struck me that most of it involved collecting data without substantive theory. Data without theory are like a baby without a parent: their life expectancy is low.”

http://www.mpib-berlin.mpg.de/volltexte/institut/dok/full/gg/GG_Surr_1998.pdf

It can rarely be said of a psychologist that everything they write is worth reading. Gigerenzer is one such psychologist. He writes in plain English (presumably his second language) and understands his material so thoroughly that he can explain it simply, the sign of an intelligent and honest teacher. This straightforward approach means that you can follow this heuristic to make you smart: if you cannot understand him first time round, it is worth reading him several times until you do. With lesser writers, if you cannot understand them first time, turn elsewhere.

Gigerenzer's lament is about the paucity of solid theories in psychology. He laments that there are only surrogates: one word explanations (vague, unspecified references to something like “similarity” offered without any definition or metrics, even when such things are available); redescriptions (such as opium making you sleepy because of its dormative properties); muddy dichotomies (pointless battles between overlapping categories) and data fitting (exquisite mathematical models which re-describe the findings, but cannot explain them).

In defence of the beleaguered dabblers in mental philosophy, it could be argued one has to start somewhere. Data fitting may at least show you where the main currents are in the stream, even if one has no theory of fluid dynamics to explain why there should be currents in the first place.

So, where next? No good calling for more pointless theories and grand delusional castles in the sky. Perhaps we should concentrate on some basic problems.

For example, what makes problems difficult? To make that a little easier, what makes one simple problem slightly more difficult than another simple problem? Other than it having the quality of being slightly more difficult?

No time limit.

Feminism and rape: intelligent?

 

It may or may not be a coincidence that the considerable increase in the use of the word “rape” coincides with the very rapid rise in the use of “feminism” in the 1970’s. Whole books have been written on the basis of such slender associations, though admittedly mostly in the behavioural sciences.

As previously discussed, behavioural scientists are under no compunction to choose their variables according to established principles, after sustained perusal of a century of published literature, or Delphic consultation with knowledgeable colleagues. The best researchers attempt one or all of those preliminary steps, but they are not compulsory.

So, to examine what may be a chance correlation of frequency patterns it is wise to try other possible words. A reader has kindly suggested “virile” but as my reply (below) shows, this does not work. Neither do the following words, chosen by a most distinguished person, to whom I had described the problem over breakfast. “Nuclear” “culture” “ethnic” “political” “economic” “black power” and “terrorist” all fail to show the same pattern, although they undoubtedly were frequent words at the time (much more frequent than our words in question).

One word shows a clear association, though as a mirror image: “intelligent”.  As it goes down and down, rape goes up, until the two words, the formerly common “intelligent” and the formerly rare “rape” touch each other in 1997. The conclusion is clear and irrefutable (I am using behavioural science talk ironically): as we ceased to regard people as being intelligent there was a rise in rape, and feminism. Perhaps we were all too aware of dysgenic trends and, looking at people’s behaviour, the word did not spring to mind so easily.

Spooky.

 

Men and women, feminism and rape

 

Can anything be deduced from books? Reading them is pretty tedious. It is hard to detect anything about social processes from the scribbles of a single person. Books have plots, story lines and other nonsense. However, if one were to take all books and conduct a meta-analysis at the level of the words alone, one might detect a bigger picture.

Consider the simple words: “men” and “women”. The relative frequency of those words may reveal the importance of these items in the popular mind, or at least the minds of readers, probably the slightly brighter, richer and supposedly more refined of the citizenry.

Sure enough, in the 1800’s men (0.0753) were discussed almost 8 times more frequently than women (0.0079). Not until 1983 did they achieve parity of mention at 0.032, and after a brief surge in which women were ahead they are at a rough parity again. Seekers of refinement should be gratified to learn that “lady” and “gentleman” showed no such disparity. Although more common two centuries ago, those genteel words are paired together, with lady (0.004) ahead of gentleman (0.002) as they should be.

The somewhat colder terms “female” (0.0048) and “male” (0.0028) are much less frequent than man and woman, and despite small differences effectively can be ignored.

Finally we get down to the very small numbers for words which are infrequently used like “rape” which was rarely used in the 1800s (0.00027), and then became slightly more rare till the 1970s. Feminism first appears in 1910 or thereabouts, and in the 1970 there is a sharp rise in unison in both these words, peaking in 1997 and falling thereafter. Compared with stalwart words like men and women, both these word have always been rare in books.

At this point in social science reporting everything pauses, like in the Hitchcock murder movie “Psycho” where a psychologist is called in to give The Explanation. I can only advance a few cautionary words: I do not think that rape was low throughout the last two centuries, rising only in the 1970’s for two decades and then falling back. It is notoriously difficult to collect data on rape in any era, but this looks like a consequence of openness of reporting and discussion. The fall since 1997 is less easy to explain. If openness has been achieved one would expect sustained use of the word.

I think this little analysis illustrates a larger point, which formed part of the critical evaluation of “The Spirit Level: Why More Equal Societies Almost Always Do Better”. By choosing what to measure one can frame and potentially distort the search for causes. For example, the word “dominance” rises in a similar way, but the match is not so good with rape. “Masculine” also rises, but is not so good a fit. I have tried a bunch of others which might conceivably be informative: “manly, seductive, powerful, arrogant, forceful, brutal, entitled” and they are all a poor fit. The match between feminism and rape remains strong.

At the moment it seems that, as is often the case in social science musings, “more research is needed”.