Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
47 views

Lecture 2 Additional Notes Introduction To Statistical Modelling

This document introduces statistical modeling. It discusses that the choice of model depends on the nature of the data, question being asked, response variable, and explanatory variables. The key is to identify whether variables are continuous or categorical, and the type of response variable. This will lead to an appropriate statistical model, such as logistic regression for binary responses, Poisson regression for counts, normal regression for continuous variables, or ANOVA/ANCOVA for categorical explanatory variables. Models are also chosen based on causation between variables.

Uploaded by

Victor Mlongoti
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
47 views

Lecture 2 Additional Notes Introduction To Statistical Modelling

This document introduces statistical modeling. It discusses that the choice of model depends on the nature of the data, question being asked, response variable, and explanatory variables. The key is to identify whether variables are continuous or categorical, and the type of response variable. This will lead to an appropriate statistical model, such as logistic regression for binary responses, Poisson regression for counts, normal regression for continuous variables, or ANOVA/ANCOVA for categorical explanatory variables. Models are also chosen based on causation between variables.

Uploaded by

Victor Mlongoti
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

BSC 311 Design and Analysis of Experiments Lecture 2

Introduction to Statistical modelling

Introduction

The choice of a good model depends on the following:

 The nature of your data.


 The question you are trying to answer.

The key is to understand the following:

 The nature of your response variable


 The nature of your explanatory variables

In the end, you need to at least answer the following questions:

 Which of your variables is the response variables?


 Which of your variables are the explanatory variables?
 Are the explanatory variables continuous or categorical or a mixture of both?
 What kind of response variable do you have: is it a continuous measurement, a count, a
proportion, a time at death, or a categorical?

These simple keys will lead you to the appropriate statistical model.

A. Models based on response variables

S# Type of response variable Type of model


1 Binary Dichotomous logistic regression
2 Multinomial Polytomous logistic regression
3 Poisson Poisson regression
4 Continuous Normal regression
Analysis of variance (ANOVA)
Analysis of covariance (ANCOVA)
5 Time to an event Survival analysis

1
BSC 311 Design and Analysis of Experiments Lecture 2

B. Models based on explanatory variables

S# Type of explanatory Type of model


variables
1 All explanatory variables Regression
are continuous
2 All explanatory variables Analysis of Variance (ANOVA)
are categorical
3 A mixture of continuous and Analysis of Covariance (ANCOVA)
categorical explanatory
variables

C. Models based on causation

S# Type of explanatory Type of model


variables
1 Clear cause and effect Regression methods
between response variable
and explanatory variables
2 No clear cause and effect Correlation analysis if both response
between response variable and explanatory variables are
and explanatory variables continuous
Log-linear models if the response
variable is a count

You might also like