Probability and significance (stat tables, cv, Type1/2) Flashcards
interferential statistics
Rather than summarising what our data shows, inferential or statistical testing allows us to conclude whether any difference or relationship found is statistically significant. This involves selecting an appropriate statistical test which allows researchers to determine the likelihood that their results have occurred due to chance. At the end of the statistical testing process, the researcher will support one of their hypotheses and reject the other.
There are various statistical tests that we can use to determine whether any difference or relationship found is statistically significant.
A statistical test is just a set of fairly simple steps – you need to know when to use the statistical tests and how to draw a conclusion from them.
Therefore, the process of statistical testing tells us whether…
*The difference or relationship between variables is statistically significant (in which case, the experimental/alternative hypothesis will be supported and the null hypothesis rejected).
*The difference or relationship between variables is due to chance (in which case, the experimental/ alternative hypothesis will be rejected and the null hypothesis supported).
Descriptive statistics
Having collected their data, researchers use descriptive statistics (eg mean, range) to provide a summary of the data collected.
If our results are statistically significant:
*It means that there is a low probability that the results are due to chance. This means that if we had studied everyone in the target population, we would obtain similar results to those we obtained from our sample.
If our results are not statistically significant:
*It means that there is a high probability that the results are due to chance. This means that if we had studied everyone in the target population, we would obtain different results to those we obtained from our sample.
What is significance?
Significance is a statistical term indicating that the research findings are sufficiently strong to enable a researcher to reject the null hypothesis and accept the alternate hypothesis.
What is probability
is a numerical measure of the likelihood that certain events will
significance level
Researchers need to decide on a measure of probability before they can determine whether their results are statistically significant
most commonly selected significance level
in Psychology is 5% (or 0.05). A researcher who chooses a 5% level of significance and finds that their results are statistically significant will conclude that the likelihood of their results coming about through chance is equal to or less than 5%.
more stringent level of significance
1% (or 0.01). A researcher who chooses a 1% level of significance and finds that their results are statistically significant will conclude that the likelihood of their results coming about through chance is equal to or less than 1%.
more lenient level of significance
10% (or 0.1). A researcher who chooses a 10% level of significance and finds that their results are statistically significant will conclude that the likelihood of their results coming about through chance is equal to or less than 10%.
What causes type 1 and type 2 errors
Choosing a significance level that is too stringent OR lenient can cause problems leading to what are known as type 1 and type 2 errors
What is the risk of making a type 1 error
the same as the significance level selected. For example, if you select the 5% level of significance, the risk of making a type 1 error is 5%.
Type 1 error
An error of optimists
Rejecting the null hypothesis when there is a good possibility that the results were due to chance.
This is often caused by using a significance level that is too lenient eg. p≤0.10.
Researcher has been too lenient.
Type 2 error
An error of pessimists
Supporting the null hypothesis when there is a good possibility that the results were significant.
This is often caused by using a significance level that is too stringent eg. p≤0.01.
Researcher has been too cautious.
Why is the 5% Level of Significance the generally accepted level
This is the perfect balance between making a type one and a type two error. It would be too lenient to use a 10% level and therefore make a type one error. However, it might be too stringent to always use the 1% level and make a type two error. Thus, the 5% level is the perfect compromise.
