Stat of the Week Competition: March 17-23 2012

Nominations

• 

  Ksenia

  Statistic: Demand-fed babies win IQ race – study

  “The study shows babies who are fed when hungry – with breast milk or formula – achieve higher scores in tests at ages 5,7,11 and 14 and that by age 8 they have an IQ four to five points higher.

  Dr Maria Iacovou, from the ISER, who led the research said: “The difference between schedule and demand-fed babies is found both in breast fed and in bottle fed babies.”

  “The difference in IQ levels of around four to five points, though statistically highly significant, would not make a child at the bottom of the class move to the top. But it would be noticeable.”

  “But mothers who keep to scheduled feeding times score better on wellbeing measures and report feeling more confident and less tearful.”

  This study was carried out by the Institute for Social and Economic research at the University of Essex and Oxford University.

  Sample size was 10,419 children born in the UK in the early 1990s and the study took into account a range of background factors such as parental educational levels, family income, a child’s sex and age, maternal health and parenting styles.

  The research compared babies fed to a schedule at 4 weeks old with those whose mothers tried but did not manage to feed to a schedule, and with those who were fed on demand.

  The children of mothers who had tried but failed to feed to a schedule were found to have similar higher levels of attainment in standard assessment task tests and IQ scores as demand-fed babies.

  Iacovou said: “This is significant because the mothers who tried but did not manage to feed to a schedule are similar to schedule-feeding mothers.

  “It seems that it is actually having been fed to a schedule, rather than having the type of mother who attempted to feed to a schedule – successfully or not – which makes the difference.”
  Source: Nz Herald
  Date: 19/03/2012

  I thought this statistic was really interesting and the study was well done. Child rearing is a hot topic of debate with a range of often conflicting approaches on how to best raise children. This provides interesting evidence rather than opinion, into one effect 2 different methods of feeding babies have on children later on.

  Firstly I’d like to briefly discuss why I think this study’s findings are reliable. As you can see the sample size of 10,419 was a large one, thereby it should minimize sampling error. Although we don’t know if it was a randomly selected sample by the sheer size of it I would suspect that it would be representative of a wide variety of the population.

  Another important point is they looked at BOTH milk fed and formula fed babies and saw that this did not affect results. I also like that they took note of other possible confounding variables such as income, parents’ education, sex and age of the child. By recognizing and accounting for these variables in their analysis we can be more confident that their results are not being produced by a confounding variable other than frequency of feeding. We can never be 100% certain but it definitely helps that they are narrowing confounding variables down so what we should be seeing is the effect the different methods of feeding are having.

  So we can tell straight away from their method that their results aren’t different for boys or girls, or dependent on family income or parents’ education.

  I think it was also clever of them to group mothers into fed on demand, fed to schedule and tried but failed to feed to schedule to try and ascertain if it was the type of feeding or the type of mother that is related to the later IQ results.

  Another advantage of this study is it is longitudinal ( so the kids were tested at ages 5,7,11,14) so we can see that it seems to be a continuing trend rather than a snapshot at one point in time.

  To conclude this study in my opinion is well designed, and so I can take their findings seriously. It is fascinating the effect that this study seems to be suggesting the type of feeding of a baby has on a child’s later IQ. Though it is perhaps not surprising to expect that what goes on in the early stages of human development, to have a subsequent impact later on.

  The study’s result suggests significant implications in terms of the sociological and psychological debate of the frequency of feeding times. Especially in light of the subtle reference/finding that what might be more beneficial for the child is perhaps harder on the mother.

  More studies of this type should be used in providing information for new parents and parenting books rather than the current trend of opinion pieces being sold and published as ‘expert’ advice.

  12 years ago
• 

  Eric Crampton

  Statistic: NZ downloaders simply shift tactics: after the new copyright law came into force, P2P traffic dropped by 75% but secure tunnelling and remote-access protocol more than doubled.
  Source: National Business Review
  Date: 19 March 2012

  Impossible to draw the headlined conclusion from the stats presented as we know nothing of base rates. How prevalent were seedboxes prior to the law change? Total downloading could be up, down, or constant. We just don’t know.

  http://offsettingbehaviour.blogspot.com/2012/03/base-rates-and-piracy.html

  12 years ago
• 

  Pierre

  Statistic: According to research, one in three women in New Zealand will experience physical intimate partner violence in their lifetime. There is an incident reported to the police every 7 minutes and only 18% of cases are reported. I don’t think it’s an overstatement to suggest this is an epidemic, and that we should seriously consider all of our social legislation through a lens of what it might mean for women in violent relationships.
  Source: Frog Blog
  Date: 20/03/2012

  According to this there are 417,000 instances of physical intimate partner violence every year. With a population of around 4 million, and half women, this means that up to 40% (bad stat alert!) could be affected every year.

  There has been discussion on the basis for these stats in the past, but it is worth bringing it out there. Womens Refuge are notorious for it.

  Violence is wrong, unacceptable and not tolerated in civilised society, but the ends do not justify the terrible stats on the way.

  12 years ago
• 

  Luis Apiolaza

  Statistic: Radio NZ News reports that [n]ew figures released to Radio New Zealand show the proportion of exports to China hit with tariffs reached a record high in 2011.
  Source: Radio New Zealand News
  Date: 20 March 2012

  It is not surprising that we hit a record high in tariffs as we also hit a record high in exports and, as the Annex 2: Special Agricultural Safeguard Measures states, we have to pay tariffs when we go over quantity thresholds.

  As for our Trade Minister confronting China about the tariffs, we did sign the dotted line in the trade agreement. I assume that we also did read the terms.

  The news should simply state “Dairy exports to China hit a record high”.

  12 years ago
• 

  Mark B

  Statistic: Demand-fed babies win IQ race study
  Source: NZ Herald
  Date: March 19

  I nominate the same Herald article on routine vs demand feeding as an earlier commentator, but for different reasons. The article’s claims are fairly sensational: demand-fed babies are more intelligent than routine-fed babies. Unfortunately, despite reporting these remarkable findings, the article doesn’t give any clue as to the location of the actual paper, which can be found via Google in the European Journal of Public Health. The article seems a fair representation of the abstract of this paper, but the problem is the abstract misses some fairly key information. The routine-fed babies made up only 7% of the cohort and were quite different from the demand-fed babies: they were more than twice as likely to have been admitted to a special care baby unit, weighed less than other babies at birth, and were breastfed for a much shorter duration. Not only that, the mothers of the routine-fed babies smoked more, had worse health prior to and during the pregnancy, had higher scores on a test for depression, had lower educational achievement, and were of lower socioeconomic status. It seems hardly surprising that there were differences in educational achievement between the groups in later years given all these differences. The authors say they controlled for these differences in their statistic models, but make no mention of the difficulty in controlling for confounding. Or the failure of findings of observational studies like this to be replicated in clinical trials, as has been highlighted on Statschat before.

  Are the findings likely to be true? No. It is far more likely that the differences in IQ are due to factors other than demand feeding. The abstract is not a fair representation of the findings of the paper, and so while the article may be a reasonable representation of the abstract, it is rather misleading. Hopefully, these findings won’t be used by anyone as the basis for making a decision about routine or demand feeding. I think the media should do a much better job than this, especially when researchers make sensational claims.

  12 years ago
• 

  Jennifer Bramwell

  Statistic: While these certainly aren’t guaranteed (please don’t cite us in your divorce petition), here are some things science says will make you less likely to get divorced.

  1. Quit Smoking

  A study published in 2010 found that if only one partner smoked, it caused more marital problems than differing religions, different backgrounds, even different plans for having children.

  Couples are an astonishing 76-95% more likely to get divorced if only one of them smokes. The amount increases when the wife is the partner with the habit. While couples who both smoke have it a bit better, a 1998 study found they are still 53% more likely than non-smoking couples to end their marriage.

  2. Take Up Optometry

  Analysis of census data from 2000 found some professions seem to be almost divorce-proof. As one might expect, religious workers and clergy had some of the lowest divorce rates, but there were some slightly less obvious professions in the bottom ten. These included optometrists, shuttle car drivers and transit police, with optometrists clocking in with a ridiculously low 4% chance of getting divorced. Other low risk professions included farming and nuclear engineers.

  If you think your job as a massage therapist, animal trainer, or mathematician puts you in the clear, think again. You’re in the top ten most-likely-to-get-divorced jobs.

  3. Share Chores

  Proponents of traditional gender roles in marriage often point to the fact that divorce rates increased as more women went into the workforce. But a study from the London School of Economics recently found that the stress on a marriage due to the wife working is completely offset when husbands contribute more to housework, childcare, and shopping. In turn, those couples are more likely to stay together.

  The study found that in households split along traditional gender lines, with the wife staying home and the husband contributing nothing to the housework, the divorce rate was slightly higher than when both partners worked and contributed roughly equally to the housework. When both had jobs and the husband made a “minimal contribution” to the housework, the risk of divorce almost doubled.

  4. Live in a Blue State

  You might expect divorce rates to be highest in states with more liberal residents, but you’d be wrong. Even though it was the solidly blue California that pioneered “no-fault” divorce in 1969, a 2009 census report revealed that residents of more conservative states are more likely to get divorced than their more liberal counterparts.

  The Census Bureau explained some of the possible reasons behind the trend. One, in states in the South and West of America, residents tend to marry younger than those in the Northeast, which in turn more often leads to eventual marital discord. Two, these states also have a larger population of immigrants, and the loss of a supportive familial and social network can put a strain on many immigrant marriages, resulting in higher divorce rates.

  5. Hang Out With People Who Aren’t Divorced

  But surely living in a red state isn’t causing people to get divorced. It’s obviously those other factors that contribute to the larger numbers, right?

  That’s true, but even just knowing people who are getting divorced makes you more likely to do so yourself. Studies on “social contagion” have shown that if you have a divorced sibling, you are 22% more likely to get divorced. But it isn’t just family members who affect us; divorces between friends and even friends of friends up your chances of ending your own marriage. Therefore living in a state with higher divorce rates, even if you waited until you were older to marry, still affects your chances of staying together.

  6. Marry Someone You Met at School

  Last year, the dating site eHarmony conducted the largest study ever into whether couples who met in certain places were more or less likely to get divorced. They looked at the expected number of divorces for couples who met at places like church, school, work, bars, and dating sites, and compared them to how many actual divorces occurred. While in most cases the number of actual divorces was almost exactly what was expected, the biggest difference was couples who met at school. According to this study at least, if you met your spouse in high school, college, or grad school, you are 41% less likely to get divorced than the statistics predict. The other major difference was couples who met in bars, who were 24% more likely to get divorced than expected.

  7. Have Sons

  A study of over 3 million couples found that having even one daughter increases a couple’s chances of divorce, while have sons lessens them. A first born daughter makes you 5% more likely to split up, while three girls increase it by 10%. A 2007 report stated that in any given year, 52,000 first born daughters under 12 years of age would still have an in-residence father if they had been born boys.

  There are a variety of theories on this. One is that men are more invested in raising sons and are therefore more likely to stick around. Alternatively, women may be more willing to leave bad marriages if they have daughters, to avoid modeling them as acceptable for their girls. This second theory perhaps makes more sense since an estimated 73% of divorces are initiated by the wife.

  Read the full text here: http://www.mentalfloss.com/blogs/archives/120595#ixzz1pz8uYr00
  –brought to you by mental_floss!
  Source: Mental Floss
  Date: March 18, 2012

  According to this piece of statistics if I do my phd in nuclear science instead of mathematics, then my chance of divorce will drastically decrease. Maybe I shall show this to some Stats 10x students and check their understanding of Chapter 1.

  12 years ago