Lecture 8 - Factorial Designs

# Lecture 8 - Factorial Designs - Factorial Designs I Basics...

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Factorial Designs I. Basics Factorial Design – any experiment with more than one independent variable ­ Use when you think there will be an interaction between your IV’s ­ Example 1: Does caffeine induce a degree of sobriety? ­ Interaction of caffeine and alcohol I. cont. Example 2: What is more important for operant conditioning, quality or quantity of reinforcement? – Interaction between quality and quantity. II. Interpreting Results Three outcomes to look at: 1. Interaction between IV1 and IV2* 2. Main Effect of IV1 3. Main Effect of IV2 *Most important – this is why you did expt in first place III. Example 1 Hyp: Caffeine will somewhat improve performance on a math test after subjects have consumed alcohol. IV1 = Alcohol (2 levels) – 0 drinks vs. 3 drinks IV2 = Caffeine (3 levels) – 0 cups of coffee, 1 cup of coffee, 3 cups of coffee III. Cont. DV = number of correct responses on a timed math test (20 questions) ­ Between subjects III. Cont. Data – 0 cups 0 17 3 8 12.5 1 cup 17 11 14 3 cups 17 17 13 10.67 15 III. Cont. Question – How do you tell if there is an interaction without using statistics? Answer – graph the data with a line graph If the lines are non­parallel, there is probably an interaction. 20 18 Number of Correct Answers 16 14 12 10 8 6 4 2 0 0 Cups 0 Drinks 3 Drinks 1 Cup 3 Cups Amount of Coffee III. Cont. Results 1. Interaction: Yes, there is an interaction. When no alcohol was consumed, number of cups of coffee had no effect on correct answers. However, when alcohol had been consumed, correct answers increased with increasing amounts of coffee. III. Cont. Results cont. 1. Main effect coffee – As amount of coffee increased so did number of correct answers. 2. Main effect alcohol – Subjects who consumed alcohol got fewer questions correct than those who didn’t. IV. Example 2 Hyp: Quality of reinforcement will be a more important determinant of a rat’s operant responding than quantity of reinforcement. IV1 – type (quality) of liquid solution (3 levels) – normal, sweet, acidic IV2 – amount (quantity) of liquid solution (3 levels) – small, medium, large IV. Cont. DV – Bar presses per minute Between subjects IV. Cont. Data – Acid S 10 M 15 L 20 Normal 18 31 35 Sweet 36 44 58 ­ Data taken from Hutt (1954) IV. cont S M L Acid 10 15 20 15 Normal 18 31 35 28 Sweet 36 21.3 44 30 58 37.67 46 70 60 Small Medium Large Number of Bar Presses 50 40 30 20 10 0 Acid Normal Sweet Type of Reinforcement IV. cont Results 1. Interaction – There is no interaction between quantity and quality of the liquid reinforcement. 2. Main Effect of Quality – Bar pressing increased as quality of reinforcement increased from acidic, to normal, to sweet. IV. Cont. 3. Main Effect of Quantity – Bar pressing increased as amount of liquid solution delivered increased. ...
View Full Document

## This note was uploaded on 02/17/2011 for the course PYSC 226 taught by Professor Levens during the Fall '09 term at South Carolina.

Ask a homework question - tutors are online