Overview of Key Points related to Cross-tabulation and Survey Research
Survey research involves collecting information (via questioning) from a group of people (population sample) so that you can describe something about the population. Surveys are meant to describe characteristics of a population (Fraenkel, Wallen, and Hyun, 2012).
Two types of surveys are common: cross-sectional and longitudinal. Cross-sectional surveys are given at one point in time and are from a predetermined population sample. Longitudinal surveys are collected at various points in time and study changes over time. Three samples of longitudinal studies are trend studies (different samples over time), cohort studies (different people from same sample over time), and panel studies (same sample over time) (Fraenkel, Wallen, and Hyun, 2012).
There are distinct steps in survey research. Like all research processes, survey research should begin by identifying the problem. It’s also recommended to write your biggest survey questions first (broad to narrow), but place the most difficult (largest) questions toward the end of the survey itself. You also need to identify the target population. Identifying a data collection tool is the next step. There are advantages and disadvantages to the various data collection formats for survey research. Highly recommended approaches are direct administration (higher participation rate and low time and cost) and online surveys (fast, easy data collection and analysis). Researchers need to create the instrument (closed and open-ended questions, peer review for formatting issues, pre-test of the instrument, cover letter, and training for the interviewers). Survey questions should be unambiguous, short, written in common language and free of bias. When reporting survey results, the researcher mush include the total sample size, the percentage of return, the per item return rate, and the percentage of responses per item.
Cross-tabulation, or contingency table analysis, is a statistical tool that is used to analyze categorical data. Cross-tabulation shows how two variables are related to one another. Let’s say there was a survey given to teachers about the quality of the district’s science curriculum. You could create a contingency table to easily display the survey questions and the frequency results. A non-parametric chi-square test would allow you to run a statistical analysis to see if there were any relationships (Huck, 2012). If the variables have no relationship, then the researcher is unable to reject the null hypothesis (and the research hypothesis must be rejected). If the variables have a relationship, then the chi-square test is statistically significant and the researcher can reject the null hypothesis.
Connections
I would like to consider using cross-tabulations to analyze the relationship between teachers’ scores on the observation and practice section of their end of year evaluation to their students’ achievement test scores on last year’s PSSA/Keystone state examinations. Cross-tab analysis makes it easy to see the graphic distribution of frequency in response. In addition, the actual chi-square test is relatively easy to understand how to conduct, even without a statistics software program. I am ready to complete last year’s staff evaluations at my new building. It will be interesting to see if there is a relationship between how students do on the exam and how well teachers were rated at their job in each of the four categories of teaching.
Most Valuable Points
The most valuable points relate to conducting a chi-square test. I also appreciated seeing examples of contingency tables and realizing that these concepts are not too difficult to understand. Additionally, the information about survey research is clear. It would be easy to use these ideas to create a survey instrument and use cross-tabulation as a data analysis tool.
References
Huck, S.W. (2012). Reading statistics and research (6th edition). Boston, MA: Pearson.

Fraaenkel, J. R., Wallen, N. E., & Hyun, H. H. (2012). How to design and evaluate
research in education (8th edition). New York, NY: McGraw-Hill Companies, Inc.