Validity
Researchers are trying to see if scores on a new and improved IQ test correlate with the previous IQ test.
Concurrent validity
A group of people take an IQ test today and then again next week. They look to see if their scores correlate with each other.
Test re-test reliability
A survey about depressive symptoms
Self-report measures
Your political affiliation, i.e. Republican, Democrat, Independent, Other
Nominal
Items on an IQ (intelligence quotient) test seem to measure intelligence
Face validity
Two researchers observe children on the playground to look for aggressive behavior and then compare scores.
Inter-rater reliability
Behavioral measures
A likert scale with answer options of: strongly disagree, disagree, neutral, agree, strongly agree
A new IQ test has several parts to it in order to try to tap into different areas of intelligence. Each part should correlate with each other.
Convergent validity
A statistics teacher wants to make sure questions on her exam are consistently measuring student knowledge, so she splits the exam questions into 2 groups then compares the students' scores on one half with the scores on the other half.
Split half reliability
Physiological measures
Previous research has shown that students with higher levels of stress are more likely to engage in academic cheating. You set up a study where participants fill out a survey that asks them how stressed they are on average on a scale of 1 to 5 and how many times they have cheated in their academic classes. What are the two variables and their scales of measurement?
1. Stress levels
- Interval
2. Academic cheating
- Ratio
Previous research shows that intelligence correlates with academic achievement and income. If my test was able to predict income based on intelligence, then my test would be good to use.
Predictive validity
Two researchers observe a male chimpanzee to track their eating patterns and report back to each other.
Inter-rater reliability
A PET scan
Physiological measures
Research has shown that people who drink alcohol perform worse on cognitive tasks. You set up and experiment where you give half of the participants 2 beers and the other half 2 non-alcoholic beers (but you tell them they are real beers). Then you give all the participants a list of words to memorize. After 10 minutes, they are asked to write down as many words as they can remember from the list. You mark down the number of words they remember correctly. What are the two variables and their scales of measurement?
1. Alcoholic vs. non-alcoholic beer
- Nominal
2. Number of words remembered
-Ratio
Intelligence and entrepreneurial skills are not related even though we probably think that entrepreneurs are very intelligent. Not all intelligent people would make good entrepreneurs, though. There should be a weak relationship between IQ scores and entrepreneur skill scores because they are distinctly different constructs.
Divergent validity
A team of researchers has a group of emerging adults fill out a survey about their self-esteem and then again ten years later.
Test-retest reliability
Behavioral measures
Previous research has suggested that people age 25 and older are more likely to consider themselves an adult than people under the age of 25. You place people into 2 categories: 25 and older or under the age of 25. You also ask people whether they consider themselves adults and they can answer: "yes," "no," or "in some ways yes and in some ways no." What are the two variables and their scales of measurement?
1. Two age groups
- Ordinal
2. Whether they considered themselves adults or not
- Ordinal