I could say about this finding -
as well
as the main finding of no differences between singles and couples in attachment issues - that the
sample was
unrepresentative (the singles were recruited by newspaper ads and the couples were recommended by the singles) and so we need to be cautious.
As Tol explains, the Cook et al paper used an
unrepresentative sample, can't be replicated, and leaves out many useful papers.
As with the land / ocean bias, a problem arises when our
sample contains
unrepresentative proportions of different strata from the population, where the strata behave differently from one another.
As the Australian blogger Joanne Nova summarised Tol's findings, John Cook of the University of Queensland and his team used an unrepresentative sample, left out much useful data, used biased observers who disagreed with the authors of the papers they were classifying nearly two - thirds of the time, and collected and analysed the data in such a way as to allow the authors to adjust their preliminary conclusions as they went along, a scientific no - no if ever there was on
As the Australian blogger Joanne Nova summarised Tol's findings, John Cook of the University of Queensland and his team used an
unrepresentative sample, left out much useful data, used biased observers who disagreed with the authors of the papers they were classifying nearly two - thirds of the time, and collected and analysed the data in such a way
as to allow the authors to adjust their preliminary conclusions as they went along, a scientific no - no if ever there was on
as to allow the authors to adjust their preliminary conclusions
as they went along, a scientific no - no if ever there was on
as they went along, a scientific no - no if ever there was one.