What to Measure and How

Measurement generally refers to the assigning of numbers to indicate different values of variables.  

If your research involved determining if there was a relationship between height and weight of dogs, it would make sense to measure the height and weight of dogs using a scale.    But if you were attempting to determine if there is a correlation between self-esteem and achievement of fifth grade girls, what scale could you use?  Would you look at grades, scores on a IQ test, how well each girl created a project?  How can such abstract concepts like intelligence or self-esteem be measured?

• What to measure (powerpoint)
• Reliability and Validity of Measures
• Pre-existing Measures
  • Achievement Measures 
    • Iowa Test of Basic Skills
    • Ohio Graduation Test
    • Prairie State Achievement Exam
    • Stanford Achievement Test
    • Wechsler Individual Achievement Test
    • Wide Range Achievement Test
    • Woodcock Johnson Psychoeducational Battery
  • Psycho-social/Health Measures
    • Chartered Society of Physiotherapy Database of Measures
  • Measures of Personality and Social Psychological Attitudes 
  • Communication Research Measures 
  • Happiness Measures

Collecting Data

Once you have decided what types of data you need for your study, you can determine whether your data can be gathered from existing sources/databases or whether you need to collect new information through other means.  Even when using existing data, it is important to know how the data was collected so that the limitations of the generalizability of results may be determined and the proper analyses may be performed. 

The researcher should be able to select and defend an appropriate method of data collection.

• Collecting Data/Sampling Error
• Survey Design
  • The Survey System
  • Survey Monkey Design Software  
• Sample Size  Formula and Calculator
• Sampling (Probability vs. Non Probability Sampling)

Summarizing and Displaying Data

Measures of Central Tendency indicate what is typical of the average subject

Measures of Variance indicate the distribution of the data around the center.

• Standard Deviation and Variation

Correlation refers to the degree to which two variable move in sync with one another.

• Correlation Tutorial 
• Scatter plots and Correlation

Graphical Displays are powerful tools for teaching and persuasion.  A picture is really worth a thousand words as many people understand pictures better than a lecture.

• Handling All That Data - Tools for Displaying Data 
• Statistical Graphs
• Statistical Software with Tutorials
  

Analyzing Data and Interpreting Results

Hypothesis Testing:

Hypothesis testing is the use of statistics to determine the probability that a given hypothesis is true.  

At the beginning of your research project, you no doubt started with an idea you thought might be true.  Perhaps you think that the oral polio vaccine causes HIV.  This is your hypothesis.  No matter how much data you collect, there is always a chance that someone might contract HIV due to "chance" (all other reasons besides the vaccine).  So in order to gather support for your claim, you have to first nullify the idea that the phenomenon occurred due to chance.  In hypothesis testing, we look to reject or fail to reject the null hypothesis.  Rejecting the null hypothesis gives credibility to your hypothesis being true, but at this point more research would probably still be in order. Failing to reject the null hypothesis does not mean the null hypothesis is true.  It simply means that there is not enough data to prefer one hypothesis over the other.

Hypothesis Testing

From SlideFinder

• Tutorial on Hypothesis Testing
• Steps in Hypothesis Testing
• Sample Hypothesis Testing Problems Glossary of Important Terms in Hypothesis Testing Rice Virtual Lab in Statistics
• Statistical Inference Activities Using TI Software

Still confused about conducting a hypothesis test?  There are lots of software applications that can do the test for you and display the results in top notch fashion.  See some statistical analysis packages, here.