name

Third-variable problem

• Example: as the number of volunteers increases so does damage from natural

  disasters with the confounding variable being size of the natural disaster.

• Internal – the accuracy of conclusions drawn about cause and effect. A study has a

  high internal validity when strong inferences can be made that one variable caused

  changes in the other variable.

• External – the extent to which findings of a study can accurately generalize to other

  populations and settings.

Validity (internal)

• – the accuracy of conclusions drawn about cause and effect. A study has a

  high internal validity when strong inferences can be made that one variable caused

  changes in the other variable.

• the extent to which findings of a study can accurately generalize to other

  populations and settings.

Ways to test reliability

test-retest

• Having someone take an IQ Test on one day and having them take the

  same test a week later. Then comparing the scores and calculating the correlation

  coefficient to find a relationship.

• Taking a 50-question test and splitting it into two 25 question test. Then

  comparing the two to find a correlation.

• two individuals would observe waiting time and atmosphere in a

  Clinique and the two observers agree than the reliability of the two raters would be

  perfect.

Ways to test reliability

split half

• Taking a 50-question test and splitting it into two 25 question test. Then

  comparing the two to find a correlation.

• two individuals would observe waiting time and atmosphere in a

  Clinique and the two observers agree than the reliability of the two raters would be

  perfect.

• Having someone take an IQ Test on one day and having them take the

  same test a week later. Then comparing the scores and calculating the correlation

  coefficient to find a relationship.

Ways to test reliability

inter-rater

• two individuals would observe waiting time and atmosphere in a

  Clinique and the two observers agree than the reliability of the two raters would be

  perfect.

• Taking a 50-question test and splitting it into two 25 question test. Then

  comparing the two to find a correlation.

• Having someone take an IQ Test on one day and having them take the

  same test a week later. Then comparing the scores and calculating the correlation

  coefficient to find a relationship.

Ways to test validity

Concurrent

• scores on the measure are related to the criterion measured at the

  same time

• The extent to which the measurement or manipulation of a variable

  accurately represents the theoretical variable being studied. In the case of

  measurement, the measure that is used an accurate representation of the variable.

• The content of the measure is linked to the universe of content that defines

  the construct.

• Scores on the measure are related to other measures of the same

  construct.

Ways to test validity

Construct-

• The extent to which the measurement or manipulation of a variable

  accurately represents the theoretical variable being studied. In the case of

  measurement, the measure that is used an accurate representation of the variable.

• The content of the measure is linked to the universe of content that defines

  the construct.

• Scores on the measure are related to other measures of the same

  construct.

• The content of the measures appears to reflect the construct being measured.

Ways to test validity

Content-

• The content of the measure is linked to the universe of content that defines

  the construct.

• Scores on the measure are related to other measures of the same

  construct.

• The content of the measures appears to reflect the construct being measured.

• Scores on the measure are not related to other measures that are

  theoretically different.

Ways to test validity

Convergent

• Scores on the measure are related to other measures of the same

  construct.

• The content of the measures appears to reflect the construct being measured.

• The content of the measure is linked to the universe of content that defines

  the construct.

• Scores on the measure are not related to other measures that are

  theoretically different.

Ways to test validity

Face

• The content of the measures appears to reflect the construct being measured.

• Scores on the measure are not related to other measures that are

  theoretically different.

• Scores on the measure are related to other measures of the same

  construct.

• The content of the measure is linked to the universe of content that defines

  the construct.