Question

A1. With the aid of examples briefly explain the difference between fixed effect and random

effects models in experimental design.

Answer 1

Definitions:

Fixed effect model:

We have tested hypotheses about, and calculated confidence intervals for, comparisons in the effects of these articular treatment factor levels. These treatment effects are known as fixed effects since we represent them in the model as unknown constants (parameters). Models that contain only fixed effects are called fixed-effects models.

Ex. Age, Sex.

Random effect model:

The effects of the levels used in the experiment are then represented as random variables whose distributions are the distributions of values in the population. Such treatment-factor effects are called random effects, and the corresponding models are called random-effects models.

Ex. The cost of a new bike varies depending on what year it was purchased (e.g. 2001 vs. 2020).