First, in regards to the last blog posting, I'm not sure if everybody hit the nail on the head, so to speak. An example I came up with concerned outliers. Remember in the beginning of the semester when I mentioned how the average salary of the comm grad is very high, something like $60,000. This is a product of using the mean to calculate the average salary. This number is inflated because of the communication department producing extremely high salary careers such as professional athletes and actors. The median would give us a much more accurate account of something like $35,000.
Today, we discussed correlation. Correlation is a statistical test that requires two continuous variables to determine if their is a relationship between the two. For discussion, I want you to come up with your own possible correlations. Pick one variable, and detail how it would be positively correlated with a second variable, and negatively correlated with a third. Explain why you think this is so.

Tuesday, September 19

In the past two days, we have discussed many of the ground-work that will be utilized later when discussing advanced statistics. We talked about;
1.) Nonparametric and parametric data
2.) Descriptive and inferential statistics
3.) Frequency Distribution
4.) Measures of Central Tendency
5.) Graphic distribution of data
6.) Standard deviation and variance

Question for thought- Generally, when describing the data, we use the mean as our measure of central tendency. For your post, give an example where the mode or median may explain the data better. Explain the variable you are investigating, and give the circumstances why the mean may not tell you information you could obtain from the median or mode.

Tuesday, September 19

In the past two days, we have discussed many of the ground-work that will be utilized later when discussing advanced statistics. We talked about;
1.) Nonparametric and parametric data
2.) Descriptive and inferential statistics
3.) Frequency Distribution
4.) Measures of Central Tendency
5.) Graphic distribution of data
6.) Standard deviation and variance

Question for thought- Generally, when describing the data, we use the mean as our measure of central tendency. For your post, give an example where the mode or median may explain the data better. Explain the variable you are investigating, and give the circumstances why the mean may not tell you information you could obtain from the median or mode.

Tuesday, September 12

In class today, we discussed reliability and validity. Remember, reliability is a necessary but not sufficient condition to have validity of an instrument.
Reliability was discussed in;
1.) Test-retest method
2.) Split half
3.) Internal Consistency
4.) Coder reliability
5.) Reliability in interviews.
Validity was discussed in;
1.) Content (face) validity
2.) Predictive (criterion) validity
3.) Construct Validity

In class today, we discussed how GRE tests have predictive validity for GPA scores in graduate school. For discussion, list a measurement that is known to have good predictive validity. List the instrument and what it predicts.

Thursday, September 7

Fun class today ya'll, I really liked how you worked through your research questions. Today, we concentrated on levels of measurement and types of error.
1.) Nominal, Ordinal, Interval, and Ratio level data.
2.) Instrument, Application, and Random error.

For blogging this weekend, I want you to think of a variable that can be measured two different ways. Give examples of how the same variable can be measured on the nominal, ordinal, interval, or ration level.

Tuesday, September 5

Good work in class today with looking deep into research results and thinking through what went wrong in the study. Not too much was discussed today, but what we did talk about is very important. Today, we went over different types of measurement on questionnaires;
1.) Open ended questions
2.) Multiple choice questions
3.) Thurston scale
4.) Likert- type scale
5.) Rank order questions
6.) Check list
7.) Semantic differential
For today's' question for thought, I would like you to create a research question, and discuss how one type of measurement would be helpful to answer the question. In addition, I would like you to describe how one other type of measurement would not be the best measurement technique.