PNB 3RM3 Unit 3

Single Factor Design (1IV & 1DV)

Between Groups Design

Within Groups Design

Advantages

Disadvantages

1. No Carry-Over Effects for participants

  1. No need to create equivalent versions of the same task
  1. Low concern for revealing your hypothesis
  1. Lack of Equivalent Groups

When to Use

Your research question demands separate groups

  1. Uses A LOT of Resources

Your experiment is too demanding for within groups

It makes sense and you like it

Assigning Participants to Conditions

Independent groups design (Random Assignment)

Advantage: Every participant has an equal chance of placement

Disadvantage: Not COMPLETELY Equivalent (b/c laws of chance- chance one group may have one type of participant)

Solutions To Disadvantages

1. Safety in Numbers (lots of participants, 30 per IV)

2. Block Randomization

Matched Groups Design (Pretesting for a deviate, then assigning to a condition)

Advantage: Gives even groups

Disadvantages

  1. Impractical: you have to pre-test
  1. Hinting at Hypothesis - Solution = running a standardized test

Natural Groups Design : Use subject variables to assign participants

Advantage: good for testing Subject variable as IV (Ex. gender)

Advantages

  1. Require Fewer Participants to obtain the same data per condition
  1. Equivalent Groups: same participants, don't need to control for individual differences

Disadvantages

  1. High concern for figuring out hypothesis
  1. Other Advantages: Resource sparing benefits, special populations with low participant pool, only way of answering quesiton
  1. Attrition: when participants leave mid-experiment
  1. Order Effects: when order of IV conditions affect your DV

Selective attrition- very bad, people leaving due to treatment condition/experimental design

Progressive Effects: First treatment affects DV on Second treatment (Produces a CONFOUND)

Practice Effects: Performance improves as a result of previous exposure to conditions

Carry-Over Effect: Response in one condition is uniquely influenced by responses in previous conditions

Solution to order effects: Counterbalancing - changing the levels of exposure of your IV so no level has an advantage or disadvantage over other levels; makes effects extraneous rather than confounding

Exposure to Levels of IV once

Ways of Measuring a DV

Concurrent-Measures Design: testing all levels of the IV at once, then measuring the DV

Repeated-Measures Design: Testing the IV, then DV, then DV again ; Participants experience different levels of the IV in different orders

Complete Counter Balancing

Advantage:: takes into account all possible confounding factors

Disadvantages

  1. High number of orders: you need n! orders (more than 4 levels of an IV is not feasible)
  1. Not feasible: recruiting/running a number of participants over many levels

Latin Square (Partial Counterbalancing): 3x3 matrix filling in missing items of conditions for participants

Advantage: Not as extensive or time consuming as complete counterbalancing

Random Selected Orders: Create a bunch of different orders, randomly select the set of orders for the participant

Disadvantage: Doesn't COMPLETELY counterbalance

Exposure to Levels of IV 2 or more times

Block Randomization

click to edit

Reverse Counter Balancing

Disadvantage: fatiguing participants