I have been busy because I have decided to write a book on education. My subject is research-based education and does it exist? Here in Ontario we are big on research based or brain based education. I am always a bit dubious about this as there are at least three things that need considering in implementing education based on research.
The first is the reliability of the research For readers like me who don’t always remember the difference between those two important pillars of good science, reliability and validity, I will explain.
Reliability refers to whether an experiment can be done more than once and by other researchers and still get the same result. If your dog eats a tablespoon of peanut butter and then lies down and rolls over twice, can you get the same effect the next day when you feed him peanut butter? If you can, can your friend in the next city get her dog of a different age to lie down and roll over twice after he has eaten a tablespoon of peanut butter? Will it work with different breeds or only black dogs weighing more than 60 lb.? The more often replications of the experiment end up with the same result, the more likely it is to be reliable.
The second pillar of research is validity. This is not as simple a concept to explain. Validity requires that the thesis and experiment make sense i.e. they are designed using both logic and fact. The conclusions must be interpreted logically, too. As the saying goes: “data is not the plural of anecdote.”
One mistake we often make is confusing correlation and causation: a classic example is the woman who believed that it was the sign “deer crossing” that caused deer to cross at that particular spot on the road. She thought this was very dangerous as she had hit a deer three times just after passing the spot. Her solution was to move the crossing.
The mistake this woman made was to mistake the correlation of a deer crossing sign and the deer crossing the road with the sign causing the deer to cross the road. Some careful thinking about the nature of deer and their abilities would have brought the realisation that deer can’t read or follow traffic rules. The deer’s preference for crossing the road at that point was the reason the sign was posted, not the other way around.
We can laugh at this person’s logic, but how often do we see similar thinking in
World Health Organization building from the South-East, Geneva (Photo credit: Wikipedia)
our lives. Think of your friend who won’t get her child vaccinated because she believes that vaccinations kill children. The World Health Organization (WHO)
clearly outlines the faulty logic as it applies to the DPT i.e. Diphtheria, pertussis (whooping-cough) and polio:
Diphtheria-Tetanus-Pertussis (DTP) Vaccine And Sudden Infant Death Syndrome (SIDS)
One myth that won’t seem to go away is that DTP vaccine causes sudden infant death syndrome (SIDS). This belief came about because a moderate proportion of children who die of SIDS have recently been vaccinated with DTP; on the surface, this seems to point toward a causal connection. This logic is faulty however; you might as well say that eating bread causes car crashes, since most drivers who crash their cars could probably be shown to have eaten bread within the past 24 hours.
If you consider that most SIDS deaths occur during the age range when three shots of DTP are given, you would expect DTP shots to precede a fair number of SIDS deaths simply by chance. In fact, when a number of well-controlled studies were conducted during the 1980s, the investigators found, nearly unanimously, that the number of SIDS deaths temporally associated with DTP vaccination was within the range expected to occur by chance. In other words, the SIDS deaths would have occurred even if no vaccinations had been given.
In fact, in several of the studies, children who had recently received a DTP shot were less likely to get SIDS. The Institute of Medicine reported that “all controlled studies that have compared immunized versus non-immunized children have found either no association . . . or a decreased risk . . . of SIDS among immunized children” and concluded that “the evidence does not indicate a causal relation between [DTP] vaccine and SIDS.”
Looking at risk alone is not enough however – you must always look at both risks and benefits. Even one serious adverse effect in a million doses of vaccine cannot be justified if there is no benefit from the vaccination. If there were no vaccines, there would be many more cases of disease, and along with them, more serious side effects and more deaths. For example, according to an analysis of the benefit and risk of DTP immunization, if there was no immunization program in the United States, pertussis cases could increase 71-fold and deaths due to pertussis could increase four-fold. Comparing the risk from disease with the risk from the vaccines can give us an idea of the benefits we get from vaccinating our children.
A plot of SIDS rate from 1988 to 2006 (Photo credit: Wikipedia)
For more information on vaccines and childhood illnesses go to Global Vaccine Safety:
Six common misconceptions about immunization. This is a much more serious case of correlation = causation than the Deer Lady’s confusion.
The sample used should be a reasonable size and reflect the population in question. How many samples, people, classrooms or animals are needed can’t be defined theoretically, but scientists and most sensible people should know when the sample is not enough. For example if one wants to know the death rate from measles, the best sample would be all the reported cases of measles in an area or all the confirmed cases of measles. The latter would be better, unless one can safely assume that doctors are generally accurate in diagnosing measles and therefore their reports won’t skew the data.
One of the cruellest results of poor research is the myth that the MMR (mumps, measles and rubella) vaccine causes autism . The research was published in 1988 and retracted by the eminent medical journal, Lancet in 2012. Not only was the research retracted but the author was also reprimanded by Britain’s General Medical Council and stripped of the right to practice medicine in Britain.
The doctor used only 12 children for his research, taking the blood samples from children at his son’s birthday party. There were other flaws in his work; for more information see the sites below.
After his results were published in 1988, some British parents refused to get their children immunized with the MMR vaccine and the incidence of all three diseases increased. Measles is highly contagious and can lead to more serious illnesses or death. For example: one in twenty will develop pneumonia (a common cause of death from measles) and one in a thousand will develop encephalitis, putting them at risk for convulsions, deafness, mental retardation or death. By 2008 there were enough measles cases in Britain to declare it an epidemic. See web sites in the bibliography below for a discussion of MMR vaccination from at least two opposing perspectives.
Even when research is well done, there are two more pits for the unwary: drawing conclusions and applying the results correctly. If your dog rolls over repeatedly after eating peanut butter, is it to please you in order to get more peanut butter or does peanut butter put his belly in such agony that he needs to roll over more than once to relieve it? One might argue that the dog’s motivation doesn’t matter; the important thing is that the dog rolls over. It does, however, make a difference to dogs and to dog lovers. They care whether pain or delight is causing the new tricks. This is another case where the researcher will have to explore the connection between the incidents in order to do good science.
Research intent on testing the results of other studies is not glamorous and doesn’t get the headlines (or the grants, sometimes) but it is as important as the initial work. In fact, without it, we would have more drugs with disastrous side effects, more collapsing structures and poorer educational systems.
The third thing that needs considering, besides reliability and validity and the conclusions draw by the researcher, is the interpretation of research by the layman – or woman. It is easy to misunderstand research if we don’t read the work or summaries without a critical eye. I find myself increasingly wondering who did the original research, how valid and reliable it was, if the researcher had a bias towards the results and what other research has been done. Education needs good research to inform good teaching practices and teachers need to know how to read the research, question it and implement what has been learned.
I have not covered everything you need to know about scientific methods and the methods of science. My intention here is to draw attention to the layman’s need to understand scientific thinking and reflect critically on research before applying it in the field. For a more thorough analysis, go to: The Scientific Method vs. Real Science at http://www.av8n.com/physics/scientific-methods.htm. It does require some thoughtful reading but it is worth the effort.
So, I am writing a book about the relationship between education and what we really know about the brain and relevant psychology. I am still in the research stage. Instead of doing my own original research, I am reviewing other peoples’ studies to understand the results and their relevance to education.
My blog will probably have a different flavour, as it is likely to reflect my thoughts and discoveries as I learn. I hope you enjoy accompanying me on this journey.
Donna, The Deer Lady
The MMR Vaccine Discussion.
Autism-vaccine study retracted Tuesday, February 2, 2010 | 10:08 PM ET CBC News http://www.cbc.ca/news/health/story/2010/02/02/autism-mmr-lancet-wakefield.html