This is one of the fundamental building blocks of the “Null Hypothesis Significance Test” (NHST) paradigm (q.v.) and “inferential testing” (q.v.)
The null hypothesis is a model of a population and is essentially that whatever of interest you are interested in (a quantitative question) is not happening there. It really is null.
The null hypothesis is specific to the question to be explored in the data you have, so for a question about whether the mean of a set of scores has changed the null hypothesis is that in the population the mean change is zero. If the question is whether trained therapists score higher than do trainees on a test of observation after watching a video then the null hypothesis is that there is no difference between the mean scores of the two populations. If the question is whether higher scores on a measure of authenticity are correlated with scores on a measure of spontaneity the null hypothesis is that there is no correlation between the two scores in the population.
The eagle eyed reader will have noticed that I’m hammering the “in the population(s)” issue. Sorry but that’s because it’s fundamental! The null hypothesis is about the population and generally the assumption in null hypothesis testing is of an infinitely large population. Generally the population is a theoretical one: we are talking about models. This means that our tests don’t “prove” or “disprove” the model, they simply give us a rational way to “accept” or “reject” it. More detail in the entry on the NHST paradigm.
Try also #
Bootstrapping / bootstrap methods
Confidence intervals (CIs)
Type I error
Touched on in Chapter 5.
Online resources #
None yet and if there are, they will present both NHST and estimation methods with some clarification of the issues with both.
First created 20.viii.23.