Examples of Experimental Variables - in development


You need to place each of the variables and factors in your project in one of the data types below and read the accompanying notes about possible analytical methods.


Scale / Continuous data

Wide variety of experimental results:

Chemistry:- Retention times in GC and HPLC, Wavenumbers in FTIR

Bioscience:- Body fat, Lung function (volumes), Reaction time

Physical science:- Pressure, voltage, current


Scale / Discrete data



Proportions / Percentages


Percentage recovery of compound




A chi-squared analysis is a common technique, where frequencies are compared for combinations of different conditions. For chi-squared analysis the data does not need to be normal.

With sufficient counts and replications it is possible to treat the frequency as a continuous variable, possibly with a normal distribution..


Counting numbers of hairs

Contractions of pig bladder


Ordinal data


This data is non-parametric in that any value associated with it only gives a ranking to the data and does not have any further quantifiable significance.

Hence it cannot be assumed to be normally distributed, and it is necessary to use non-parametric statistics, e.g. Mann-Whitney, Wilcoxon, Kruskal-Wallis.

Under certain circumstances you may choose to demonstrate the use of parametric statistics, e.g. ANOVA, for this data, provided you make it clear in your discussion that the results may be unreliable.


Quality assessment


Nominal data


This data has no quantitative or ranking value. Usually appears as a factor variable in the experiment design.


Gender (M/F)

Smoker (Yes/No)

Fabric type