Please enable JavaScript.
Coggle requires JavaScript to display documents.
STATISTICAL TESTING, test of difference, test of correlation, nominal data…
STATISTICAL TESTING
- to decide which statistical test is used under what circumstances, there are 3 factors:
- decision 1: difference or correlation?
- consider first when deciding a statistical test, the aim/purpose of the investigation - is the researcher looking for a difference/correlation?
- should be obvious from wording in the hypothesis. 'correlation' can include investigations which are looking for an association
- decision 3: levels of measurement?
- quantitative data - can be divided into different levels of measurement & this is 3rd factor in influencing choice of statistical tests.
- nominal data - represented in form of categories hence nominal data is sometimes called categorial data
- .e.g. you ask everyone in class if they like psychology. people who say 'yes' are in one group. people who say 'no' are in another group
- nominal data = discrete in that one item can only appear in one of the categories, e.g. if you asked people to name their favourites football team their vote only appears in one category (nominal data can have more than two groups)
- ordinal data - data ordered in some way, e.g. asking everyone in class to rate how much they like psychology on scale of 1-10
- doesn't have to have equal intervals between each unit (unlike in interval data)
- ordinal data lacks precision as its based on subjective opinion rather than objective measures
- interval data - based on numerical scales that include units of equal, precisely defined size
- in this sense its 'better' than ordinal data because more detail is preserved (and ordinal data is 'better' than nominal)
- most precise & sophisticated for of data in psychology & is a necessary criterion for the use of parametric tests
- decision 2: experimental design?
- if investigation is looking for a correlation, not difference, experimental design isn't an issue
- repeated measures & matched pairs are referred to as related designs
- repeated measures design, the same ppts are used in all conditions of experiment
- in matched pairs design, ppts in each conditions aren't the same but have been 'matched' on some variable that is important for the investigation which makes them 'related.
- ppts in each conditions of an independent groups design are different so this design is unrelated
- statistical test is used to determine whether a difference or an association/correlation found in a particular investigation is statistically significant.- that is, more than could have occurred by chance.
- outcome of this has implications for whether we accept or reject the null hypothesis
-
-
-
-
-
-
nominal data = a level of measurement, data which is in separate categories
ordinal data = a level of measurement, data which is ordered in some way but the intervals between each item are unequal
interval data = data measured on a scale where the distance between each value is the same, such as when counting correct answers or using any 'public' unit of measurement
-
-
-