Tuesday, August 3 Chapters Six & Seven Educational Research
1. Nominal scale: (text pg. 133) numbers assigned to categories.
The researchers used a nominal scale in assigning “1” to males and “2” to females.
2. Ordinal scale (text pg. 133) numbers rank ordered.
The participants in the study were lined up by height with the tallest person being 1st and so on down the line.
3. Interval scale: (text pg. 134) equal intervals between numbers.
When John scores 90 and Hank scores 80 on an interval scale, we know they are the same difference as June who scores 20 and Jill who scores 30.
4. Ratio scale: (text pg. 134) numbers expressed as ratios.
When starting a scale based upon height, if you start at 0 it can be a ratio scale as there are equal intervals between inches.
5. Histogram: (text pg. 135) bar graph.
The researchers used a histogram to display the data in graph form.
6. Positively skewed: (text pg. 137) large numbers of low scores; few very high scores.
In a study of administrators’ salaries, the distribution would be positively skewed if most of the salaries were relatively low and a few were very high.
7. Standard deviation: (text pg. 139) average distance of the scores from the mean.
A standard deviation of 35 shows the reader that the scores in the study were very widespread across the board.
8. Correlation coefficient: (text pg. 142) number between -1 and +1 that indicates the direction and strength of the relationship.
When the correlation coefficient is closer to +1, the relationship between two variables being studied is strong.
9. Validity: (text pg. 144) the extent to which inferences are appropriate and meaningful.
A test for beginning teacher competency may be valid for how much those teachers know about classroom management.
10. Reliability: (text pg. 149) consistency of scores.
The more ambiguous questions are and the more tired the participants are when taking the test, the less reliable the instrument for that test proves to be.
11. Stability (test-retest reliability): (text p. 150) measured by giving the same instrument twice.
Researchers found high stability in their test of academic achievement by testing the group in June, intervening with a tutoring program, and retesting the group in December using the same instrument.
12. Equivalence: (text pg. 151) correlation of two forms of the same test.
Rather than giving the same test, researchers can find equivalence by administering gives alternate but equal tests to the same group.
13. Norm-referenced: (text. 160) interpretations that compare subjects with others.
A reference group, called the norm group, is the group that scores on the test are compared to when using a norm-referenced instrument.
14. Criterion-referenced: (text. 161) interpretations that compare subjects with a standard of performance.
The group being tested is compared to an established level of performance or skill when using a criterion-referenced instrument.
15. Semantic differential: (text pg. 169) a 7-point scale with adjective pairs as end points.
By placing adjectives on opposite ends of the scale as anchors, the researchers checked on attitudes of participants using a semantic differential.
The last two chapters (six and seven) of the Educational Research text were full of information regarding using instruments in research and whether they are reliable and valid. While it seemed to be quite overwhelming (evidenced by my list of 15 vocab words!), I feel that I am getting the hang of it. Some of the information is easy to digest such as the different types of scales. I think this is due to the activities we performed in Educational Inquiry class. By getting up and visually showing the class what a nominal, ordinal and interval scale looked like using height, the information sticks in our brains a bit more. I find this helpful in my classroom as well and it was a good reminder to try this more often. It especially helps to use something so visual like height, when we see the difference between Andrew and David as compared to Molly and Janelle.
I also liked seeing the visual on the PowerPoint of positive and negative correlation. It helped to see a program where the graph can move and we can see the scatter points. This shows very clearly how scores from participants in a study can be very close together or quite far apart, and then the impact this has on correlation of variables. I feel that it also helped us to have to “act out” correlation. It triggers a more creative part of the brain and provides a snapshot memory of correlation for me. I can now think of no correlation as people randomly doing things in the room (no relationship or very little between two variables) and positive correlation as cartwheels and clapping.