Why Create a Statistical Model?

- You need to find solutions / answers to a problem
- You could try it out for real, but that would be expensive and time-consuming
- Using a mathematical model is therefore cheaper and quicker
- You can change and analyse the variables easily (a lot easier than putting a prototype back to how it was originally, then testing it under some different conditions, anyway)
- However… models are only simplifications, and they might not work under certain conditions!

How to Create a Statistical Model

- Observe the thing you want to model, identifying what you want to predict
- Suggest a model
- Use the model to predict some stuff
- Collect data to test the predictions
- Compare the predictions with the results
- Refine the model if necessary
- GOTO 3 (or 4)

Data Definitions

- Data – Measurements, observations, facts… (the actual bits – the numbers)
- Variables – Things that are measured or observed (mass, age, score in Statistics exam…)
- Qualitative data is non-numerical data
- Quantitative data is numerical data
- Discrete data can only take certain values (integers only, for example)
- Continuous data can take any value (within a realistic range, anyway)

Classes

- Classes are groups data can be in
- They’re what you get from saying “0 to 10 here, 10 to 20 here…”, etc.
- The class boundaries are the… boundaries… of the class
- The mid-point is… the point at the middle of the group (the mean of the two boundaries)
- The class width is… the width of the class: the upper boundary minus the lower boundary

###### Related articles

- Statistical Models, Data, Classes – AS Maths Revision – Statistics (S1) (mattg99.wordpress.com)
- Averages from Charts / Tables – AS Maths Revision – Statistics (S1) (mattg99.wordpress.com)
- Measures of Location – AS Maths Revision – Statistics (S1) (mattg99.wordpress.com)
- ‘Coding’ and Measures of Dispersion – AS Maths Revision – Statistics (S1) (mattg99.wordpress.com)

Advertisements