Data doesn’t just happen.
Plan ahead for good data.
Planning is easy.
 |
|
| 1.The Idea | Chapter 1 & 2 |
| 2.The Hypothesis | Chapter 1 & 2 |
| 3.The Variables | Chapter 3 (& 10) |
| 4.The Relationship | Chapter 4 |
The Idea
You will have an idea to start with.
It may be quite casual:
Risk-takers do better in exams
The Hypothesis
Now we make this into a formal hypothesis.
The hypothesis is nearly always a statement about variability:
Some variability in exam grades might be explained by variability in risk-taking
The Variables
Finally we convert this into a statement about variables.
A variable is a way in which participants or situations vary. These are variables:
ExamGrades and RiskTaking
Different people have different exam grades and have different risk-taking.
Keep variable names short but informative.
Do not use spaces: use underscore or CamelNotation.
The Relationship
Different variables take different roles in our hypothesis: our hypothesis is a relationship between variables.
We are saying that we want to find out whether there is a relationship
RiskTaking → ExamGrades
We are saying that risk-taking might explain (in part) exam grades. We aren’t trying to explain variability in risk-taking, but we are trying to explain variability in exam grades.
| Role | Description | |||
|---|---|---|---|---|
| the variable we are trying to explain | |||
| the variable we are using to explain |
More Complex Relationships
The hypothesis can involve many more variables, if required. Here is an example:
[RiskTaking | Perfectionism | Wakefullness] → ExamGrade
This lists 3 IVs each of which will be used to explain the DV.
The Plan (part 1)
The research plan begins to take shape. We can capture this in a simple table:
| Variable Name | Role | Type | Values |
| ExamGrade | Endogenous | Interval | Mean=50, SD=15 |
| RiskTaking | Exogenous | Interval | Mean=5, SD=1 |
| etc… | |||
| etc… |
Variable Type
There are 2 fundamental types of variability between people:
people can have different qualities (eg. red-head or not)
people can have different quantities(eg. exam grades)
These correspond to two fundamental types of value:
qualities: category labels
quantities: numerical values
Where a variable has numerical values, then we can make a subtle sub-division.
All the things we use numbers for can be sorted into an order.
Some applications of numbers can be described by an average.
So we can talk about the average exam grade – it might be 60; it doesn’t make sense to talk about the average exam placing (that would be 52nd if there are 104 people taking the exam).
Categorical variable: values are category labels.
Interval variable: values are numbers that can be averaged.
Ordinal variable: values are numbers that can’t be averaged.
Generally Interval is best, Ordinal is next best and Categorical is worst.