lawson
Chapter 13 – Analysing data I: Quantitative data analysis
Information Classification: General
1
Topics covered in the chapter
How to prepare your quantitative data for analysis.
An introduction to a data analysis software package – JASP.
An introduction to the concept of inferential statistics.
The idea and meaning of the concept of statistical significance.
Information Classification: General
Using JASP
Information Classification: General
Creating your data file
Code your data.
Input the data in the following format:
Information Classification: General
Types of Data in JASP
Nominal text – variables without numerical value. (e.g. short qualitative responses).
Nominal –data that has no natural order or ranking.
Ordinal - data with an inherent order, for example we often use scales in questionnaires that might range from 1 (“not at all”) to 5 (“very much so).
Continuous – “True” numbers, with equal distance between measures, and a true zero, for example age in years or height in cm.
Information Classification: General
Importing your Data into JASP
Click on ‘open’.
Locate your file.
Open the data file – which will appear as follows:
Information Classification: General
Labeling your values
Click on a variable name (e.g. ‘level).
Label each value by replacing the label number with its meaning (e.g. 1 with ‘amateur’, 2 with ‘professional’.
Information Classification: General
Check for outliers
Click on Descriptives > Descriptive statistics.
Move the variable to be explored into the variable box.
Then under Plots, click Boxplot element > Label outliers.
Information Classification: General
Descriptive Analysis
Descriptive analysis explores a single variable at a time.
Measures of central tendency – Mean, median, mode.
Measures of Dispersion – Standard deviation, range.
In JASP – click on Descriptives, move the chosen variable into the variable box, click on Statistics.
Information Classification: General
Inferential Statistics
Assess the association between independent and dependent variables.
These may be bivariate (measuring the effect of a single independent variable upon a single dependent variable).
Or multivariate (involving more than two variables).
Information Classification: General
Parametric tests
Use interval or ratio data, or what JASP classifies as continuous data.
Assume that the data is drawn from a normally distributed population (i.e. the data is not skewed).
Has the same variance (or spread) on the variables being measured.
Information Classification: General
Non-parametric tests
Are used with ordinal or nominal data.
Do not make any assumptions about the characteristics of the sample in terms of its distribution.
Information Classification: General
Choosing a test
Information Classification: General
Interpreting a test
Tests calculate the likelihood of whether an apparent relationship or difference between two or more groups is down to chance or not.
Generates a p-value.
A p-value indicates the compatibility between your findings and the null hypothesis, the lower the p-value (which ranges from 0 to 1), the less the compatibility.
Information Classification: General
Type 1 and Type 2 errors
There is a possibility of the answer you get not reflecting the ‘true’ result.
Type I error - rejecting the null hypothesis when it is, in reality, true.
Type II error - accepting the null hypothesis when it is, in reality, false.
Information Classification: General
Common mistakes in statistical analysis
Choosing an incorrect test, often through applying parametric tests to non-parametric data.
Collecting data in the incorrect format for the appropriate test.
Misinterpreting a p-value, or deciding upon an inappropriate level of significance, and making a type I or type II error.
Testing as many variables as possible to pick up any significant results without any rationale.
Deciding upon a level of significance after analysis.
Information Classification: General
Summary
Descriptive statistics allow you to organize and summarize data.
Inferential statistics allow you to draw inferences between two or more variables.
Inferential tests will provide you with a ‘p-value’ which indicates the likelihood that any association or difference was down to chance or not.
The importance of statistical analysis lies not in the analysis itself – Instead, it is the correct interpretation of the results!
Information Classification: General