Question: Give two examples of variables that could be measured with a survey. List three general characteristics of a good survey.
Textbook:Huck, S. W. (2012). Reading statistics and research (6th ed). Boston: Pearson Education Inc. (Chapter 12).
Answer Sample #1
The universe of variables capable of being measured by a well constructed survey is virtually without limit.
One example, for discussion purposes, might be a mental health organization measuring present attitudes, opinions, or beliefs about anti-depressants in connection with designing a potential initiative aimed at neutralizing public perception, reducing discriminatory misconceptions, and increasing the likelihood that individuals struggling with depression will seek help and maintain medication regiments. By identifying the attitudes, opinions, and beliefs that create obstacles to depression care and management, a consumer outreach program specifically speaking to positively influence those attitudes, beliefs and opinions might be more accurate designed and executed and, accordingly, more effective.
Another example might be measuring anticipated future home buyer behaviors in connection with predicting housing trends, for example for purposes of government home finance program planning or a residential developer designing a new housing community and affiliated housing products aimed at first time, move up, retired, single, married couples with children, multi-generational, or other demographics of home buyers and occupants. By identifying intended home buying behaviors, participants might reveal information relevant to programs, products, and services, which can then be better designed and equipped to meet these future needs and demands in both the public-not for profit and private-for profit sector context.
The general characteristics of a good survey include, without limitation, using short and simple questions. Doing so, among other things, helps to insure that questions are understood and responses are accordingly accurate and appropriate to the question, reducing the likelihood of inaccurate, misleading responses as a result of respondents misunderstanding questions. Avoiding leading or ‘loaded’ questions will, likewise, help to insure that respondents provide their true insight, opinions, beliefs, experiences, and such, reducing the likelihood of influence by survey instrument structure or characteristics. And pilot testing the survey instrument will help enable researchers to pro-actively identify potential flaws in survey instrument structure and characteristics, affording the opportunity to correct and fine tune survey characteristics prior to investing in a full launch.
Answer Sample #2
Two types of variables that can be measured by a survey are categorical variables and continuous variables. Categorical variables can be further classified as being nominal, dichotomous, or ordinal variables. Each of these types are known as categories or levels, which are the descriptions that are given to a variable to help explain how variables should be measured, and or controlled. For example, if I would like to know the career path that my students have chosen I would ask: What is your career path? and What attracted you towards that career? The first question is the use of categories, for example: Architect, Attorney, Banker, Engineer, etc. And the second question could be a scale question. For example: On a scale from 1 to 5, what attracted you towards this career? (1 = least important, to 5 = most important) The levels could include, nature of work, salary, job growth, etc. For research purposes, the surveyor can choose which categories or levels to include and how many categories or levels there should be. The continuous variable is a variable that has an infinite number of possible values. For example, if I was taking a survey for a fire department and the fire department mandates that all fire fighters must weigh between 150 and 250 pounds, I could ask: Answer yes/no if your weight is between the departments mandated weight of 150 and 250 pounds. The weight of the fire fighter is an example of a continuous variable because the weight could take any value between 150 and 250 pounds.
I believe good survey should address a particular topic, and it should be short because long surveys are often unappealing. The questions should be simple, clear and ask for an answer on only one dimension. For example, if a survey asks “Did you like our food and service?” If the person answers “yes”, then the researcher will not know if it’s the food they like or the service, or both. Also, questions should be non-threatening. If a person taking a survey is concerned about the consequences of his or her answers, there is a possibility their answer will not be truthful. If the survey contains sensitive items, it should always state the policy on confidentiality
