Question

for stat students,

model ( linear regression, multiple regression,factorial experiments,liner model)

for each statistical method , why is the underlying statistical model important ? more than 4 reasons.

please explain in clear way , i will discuss that with my class . Thx

Answer 1

The underlying statistical model (linear regression, multiple regression,factorial experiments,liner model) is generally a combination of inferences that is based on collected data and the population understanding; used to predict information in an idealized form.

This means that a statistical model can be in the form of an equation or a visual representation of information/data based on research that's already been collected. Notice that the definition mentions the words 'idealized form'. This means that there are always exceptions to the rules.

The purpose of the underlying statistical model for the above mentioned techniques is Prediction and Explanation..

1. Prediction: What is the output for a set of input data? How a minor change in a particular type of input data may effect on the output information..

2. Explanation: How do the variable relate to each other or the other variables? How strong is the relationship between the variables? How much of the variation in the dependent variable is explained by the underlying statistical model?

Here is a list to help consolidate your understanding of what the underlying statistical model (linear regression, multiple regression,factorial experiments,liner mode) is:

• Statistical model is an approach in statistical data analysis that helps the user to discover something about a phenomenon that is assumed to exist. This approach helps explain the variability found in the data set.
• Modelling is a unifying strategy which brings together estimation and hypothesis tests under the same umbrella. Estimation is the process of generalizing the findings from one study to a target population. Hypothesis tests help in determining how complicated the statistical model needs to be.
• A modelling approach constructs a summary model that display current knowledge. The models are then 'fitted' to the data.
• All commonly used statistical procedures can be put into a general modelling framework.

This is of the form:

Data = Pattern + Residual

where variation in the observed data can be split into two components: the Pattern – systematic or 'explained' variation – and the Residual – leftover or 'unexplained' variation.