PSY 1004 Lecture Notes - Lecture 6: Construct Validity, Quasi
9 Jun 2018
School
Department
Course
Professor
Threats to Construct Validity
●Poor measurement technique
●Poor operational definitions
○Use operational definitions to get more and more specific
Threats to External Validity
●Non-representative sample
●Using one specific times
○Historical trends change
●One settings
○Using the lab setting instead of a more realistic setting
Threats to Statistical Validity
●Improper use of statistics
●Using too few participants
Other Threats
●Role Demands
○good- subject tendency
○Evaluation apprehension
●Experimenter Bias
○Control with “blind” studies
Topic 5
True Experiments
●Scientist has “complete” control over all aspects of the experiment
○Control over what, who, when, where, and how of the experiment
○Limit factors/variables that change the DV besides the IV
●Compared to Non-experimental or Quasi Experiment
●Control is a key characteristic of true experiments
○Ruling out threats to validity
●Critically thinking
○What are the threats to validity?
○How can I neutralize those threats?
●Control practices depends on the type
of true experiments
●Data is collected either:
○Within subjects- same group of participants
○Between subjects- independent group of participants
Document Summary
Use operational definitions to get more and more specific. Using the lab setting instead of a more realistic setting. Scientist has complete control over all aspects of the experiment. Control over what, who, when, where, and how of the experiment. Limit factors/variables that change the dv besides the iv. Control is a key characteristic of true experiments. Control practices depends on the type of true experiments. Each subject is used as their own control. Simplest design: control condition and experimental condition. Can use design with >two levels of independent variable. The ordinal position of a condition changes a subjects performance. Subjects experience the conditions in a variety of orders. Block randomization- order of conditions is randomized, but each subject is presented with each condition before any condition is repeated. Reverse counterbalancing (aka mirror-image / abba )- conditions are presented first in one order, then in the reverse.
