Here are the main topics in Chapter 12 of your text.
A wildly popular method of surveying is with free online systems like Zoomerang and Survey Monkey. If you are not already familiar with this online software, you will have an opportunity this week to deliver your first assignment (your research plan) using this free software. I hope you have some fun with this! Both Zoomerang and Survey Monkey provide computer assisted data collection and automatic analysis and reporting capabilities. This week’s activity will acquaint you with this functionality. More detailed analysis calls for more expensive but more robust statistical packages such as SPSS.
Telephone surveys (and open-ended questions in mail surveys) require coding of responses. Callers (and data entry clerks) need to be trained in coding to ensure consistency.
In addition to training in coding, callers need to follow protocol to avoid biasing the responses.
Callers track responses using call sheets and apply acronyms as shorthand.
The textbook sets these benchmarks for response rates to telephone and mail surveys. Are these response rates universal or situation-dependent?
Once data is coded, it is time for analysis. Bivariate analysis permits researchers to compare two variables. Univariate and multivariate analysis permits researchers to compare multiple variables. Your final assignment in this class (the Research Report Analysis) gives an example of this.
While this chapter goes into quite a lot of statistical for all its merits…your main task is to “speak the language” of research not necessarily be a statistician . (Although if you go this route I salute you!) Your main work as a public relations practitioner is to know enough of the key concepts and terms to work with a statistician to get research results that bring value to your organization. With that, we conclude the slides for week 10. Bye for now!
HPPR404 Unit 10
HPPR 404 Research and Evaluation Sherrell Steele Collecting, Analyzing, and Reporting Quantitative Data
Objectives <ul><li>How to design surveys </li></ul><ul><li>How to train interviewers </li></ul><ul><li>What are call sheets? </li></ul><ul><li>What are acceptable response rates? </li></ul><ul><li>Reporting relationships among variables </li></ul>
Survey design <ul><li>Online templates e.g. Zoomerang, Survey Monkey (free) </li></ul><ul><li>Statistical packages e.g. SPSS (costly) </li></ul><ul><li>Computer assisted data collection (automatic data analysis) </li></ul>
Coding <ul><li>Assigns a number to a response , usually on a scale (e.g. Likert) </li></ul><ul><li>Code books help to keep consistent analysis when there are several people doing coding </li></ul>
Training interviewers <ul><li>Be polite </li></ul><ul><li>Read slowly and clearly </li></ul><ul><li>Read each question consistently </li></ul><ul><li>Do not influence (bias) </li></ul>
Call sheets <ul><li>Track timing of calls, call history </li></ul><ul><li>Use code to record e.g. CM= Completed interview </li></ul>
Response rates <ul><li>The number of people who complete a survey out of a valid sample size </li></ul><ul><li>Phone surveys - 45% to 55% response rates </li></ul><ul><li>Mail surveys - 30% response rates </li></ul>
Univariate relationships <ul><li>Compares results from several groups e.g. How many male smokers attend sporting events. </li></ul><ul><li>Reported using crosstab tables see Table 12.3 and 12.4 pages 263 and 264 in your text </li></ul>
Summary and conclusion <ul><li>You do not have to be a statistician. Leave Chi squares and correlation coefficients to internal experts or external contractors. </li></ul><ul><li>Your main task is to make the outcomes understandable and acceptable to your senior managers. </li></ul>