Quantitative research for beginners
You must understand your data.
The thing I try to explain I call response variable. = Dependent variable
The rows of the observations are the variables.
The things I suspect that are doing the explaining, are called predictor variables. = Independent variables
Nominal: a value that is a number but the number does not mean anything. E.g. For male give 1 and female give a 0
Ordinal: one value is bigger or greater than the other. Weak ordinal data: E.g. 1 denotes 4-5 hours, 2 denotes 6-8 hours and so on. But we need to know the order. You don’t know the actual value, people just ticked a box e.g. Box 1 is £0-£19999 per year etc. Strong ordinal data: you don’t know the score but you know that one is bigger than the other
Interval: looks like ratio data. You got the actual value, included decimal value.
Ratio: I want your actual income with decimal point, so zero income means no income. Most detailed quantitative data. You can summarise the data, and convert it to ordinal for example. An estimate is ratio data.
Google your quantitative data technique. Find the fieldbook for spss. Check YouTube
Standard deviation can be used as measure for variability.
Factor analysis is picking the dimensions of your data, like age, education etc
Cluster analysis is putting groups of people into clusters
Index creation how do I put them together
Principal component analysis might help me?
Normalise the scores by the use of standard deviation
The regression lines becomes my new measure.
It can take as many items as you want and come up with the dimensions.
For modelling the independent variable is always in the x axis
If the Pearson’s r is zero there is no correlation
In the simple example that is a negative indirect relationship
The b1 is the gradient, steepness of the line. If b1=0 then there is no correlation between x and y. Gradient measures how much the y to go up if I increase the x by one.[Top]