gremlin/Getty Images CHAPTER 7 Survey Research J. Dennis Blessing, PhD, PA CHAPTER OVERVIEW Surveys are one of the most commonly used methods to gather information and data in the world. Surveys are used to gather information as important as the U.S. Census or whom we plan to vote for in the next election or as mundane as the type of dishwashing soap we use or our favorite movie. Surveys can reveal trends, attitudes, opinions, lifestyles, needs, expectations, knowledge/information, behaviors, demographics the list of uses is infinite. Surveys can be used in almost every aspect of our lives, and health care is no exception. A large amount of data from any number of people on a wide variety of subjects and issues can be gathered using surveys. Several formats are used to deliver and conduct surveys; they can be long or short; simple or complex. Surveys can use words, symbols, written responses, and simple or complex choices, and can collect quantitative or qualitative data. Despite its somewhat eclectic nature, survey research must still conform to the exacting requirements of design and methodology to have true value, meaning, and impact. This chapter presents the basic concepts of survey research, including design, item construction, data analysis, and conducting the research. Emphasis is placed on designing items and developing concepts that are valid to survey research in health care. A short discussion of sample size is presented. The use of surveys in research and practice environments will be included. LEARNING OBJECTIVES Discuss and evaluate the purpose of surveys. List the advantages and disadvantages of surveys. List the various types of surveys. Define verbal frequency scales and develop a verbal frequency scale. Develop survey items. Construct a survey. Discuss and identify survey analysis procedures. 91 92 Section 2 The Research Process: Design What Surveys Can and Cannot Do Surveys are used to gather many types of information: information on demographics, opinions, preferences, expectations, knowledge, and outcomes. Surveys can be conducted by mail (including email), in person, via telephone, and through the Internet. Survey samples can be very specific for selected individuals, less specific for groups of individuals, or more general for large populations. Sample subjects can be defined by almost any characteristic. Samples can be randomized or assigned, stratified or open, well defined, or ill-defined, or a sample of convenience. As shown in TABLE 7.1, surveys have several advantages for data collection; they can be a relatively easy, inexpensive, fast, and consistent research tool. As shown in TABLE 7.2, surveys also have disadvantages. One major disadvantage is difficulty in establishing causality.1 Causality involves manipulating a variable and determining the effect on another variable. This relationship is very difficult to establish using a one-time survey. However, well-designed surveys can be effectively used as preintervention and postintervention TABLE 7.1 Some General Advantages of Surveys Advantage Reason Costs Relatively low cost, but depends on survey and methodology Sample size Can vary from small to extremely large; can be targeted Issues Can gather information on single or multiple issues or topics Format Can vary: mail, Internet, interview, telephone Impact Can yield very influential results TABLE 7.2 Some General Disadvantages of Surveys Disadvantage Reason Respondent self-report There is no way to ensure individuals will respond to surveys or choose to participate. Respondent interpretation There is a risk that respondents will interpret survey items in a manner different from investigator intent. Respondent bias There is a risk that people who choose to respond may be different from the population being studied. Extrapolation to non-responders Can the response of a few be extrapolated to the population? Can the data be generalized? Item bias Is there bias in the way an item is worded (by the investigator) or interpreted (by respondents)? assessment tools in experimental and quasi- experimental designs. Challenges associated with survey research revolve around the reliability of respondent self- reporting, respondent interpretation of survey items, extrapolation of results to a population, non-respondent influence, and respondent bias. Although these challenges to survey research cannot be completely avoided, good design and item construction should and will help to limit the impact and influence such issues have on outcomes. Types of Surveys As shown in TABLE 7.3, surveys are delivered in different formats. The choice of delivery format (distribution) is part of the methodology and Chapter 7 Survey Research 93 TABLE 7.3 Comparison of Survey Methods Characteristic Mail Telephone Interview Internet Cost Moderate, depends on sample size Low to moderate, depends on local versus long distance costs Relatively high Low, least expensive Cost examples Printing, postage, paper, envelopes Telephone costs, personnel training, and expense to conduct an interview Personnel training and expense to conduct an interview Professional help with Web design Response rate Potentially low Dependent on the willingness of subjects to participate Relatively high; only willing subjects are interviewed Variable, but instant; respondents must have Internet access Investigator time Relatively limited Requires training of interviewers and/or time to conduct interviews Requires training of interviewers and/or time to conduct interviews Relatively limited Respondent time Completed as convenient Telephone time Interview time Completed as convenient Study time Relatively long; must allow for mail and response times Relatively short, but subject contact time may be long Relatively short, but subject contact time may be long Relatively short, dependent on subject time to respond Anonymity Survey form can be completely anonymous Phone number is known Person-to-person interaction Web address can be identified Sample size Can be very large Usually small to medium Limited, usually small Can be very large Respondent bias Respondent item interpretation Interviewer voice influence Interviewer influence is high Must be computer literate Investigator bias Very limited except for item construction Voice influence; interpretation of survey items Influence of voice, facial expression, body movement; interpretation of survey items Very limited except for item construction (continues) 94 Section 2 The Research Process: Design TABLE 7.3 Comparison of Survey Methods (continued) Characteristic Mail Telephone Interview Internet Other Requires mailing lists Multiple interviewers conducting a survey Multiple interviewers conducting a survey Computer access required research design. Each format has advantages and disadvantages. The research purpose(s), study population, and sample size are factors that must be considered in the decision on the survey delivery. Surveys can be conducted by mail, telephone, Internet, and interview. Most of this chapter will concentrate on Webbased surveys and those delivered by mail. The construction of the survey itself is very similar for both formats. Many Web-based surveys and paper surveys are delivered by email. Internet surveys have become the standard format for surveys. Telephone and interview surveys are briefly discussed. Planning Is Key Performing a survey investigation or study is more than thinking up some questions, putting them on paper or online. The meticulous planning required is as important for surveys as for any other type of research effort. One of the first steps is to define what is to be achieved in the study. Not much can be accomplished beyond a literature review until the study goals are conceptualized and what is to be achieved and learned are defined. The survey must have specific goals or objectives or aims that can be researched. Once the goals, objectives, and/or aims of the investigation are defined, everything that follows will reflect the research effort. This planning generates the research questions, hypotheses, or null hypotheses. Response Rates Response rates are a threat to any survey. Response rates tend to be low, particularly when used without a well-defined target population or sample. Here are some helpful guidelines for surveys of health professionals, students, and educators. An introductory statement (cover letter if using regular mail) should accompany the survey. A personal salutation should be used if the subject is known to the investigator. Otherwise, Mr., Ms., Mrs., or a professional title such as Dr. should be used. The survey instrument must have well- designed content (items) and be visually appealing. A respondent should be able to complete the survey in a short period. The longer it takes to complete the survey, the less likely that it will be completed. Some inducement or gift may be included. However, this adds to the study expense, and there is no guarantee the response rate will be increased.1 A follow-up request may spur some respondents. An email reminder may suffice. Regarding the form and format of an email survey, these additional suggestions may be helpful. Begin with interesting questions (items). Use graphics and various question-writing techniques to ease the task of reading and answering the questions (simple is good). Use capital or dark letters for readability. Limit the length of the survey. Send notices and reminders at carefully spaced times. Notice by email that a survey is coming Email with the survey attached A reminder A second reminder (optional) Chapter 7 Survey Research If an identifier is used, explain that it is for recordkeeping only and that the respondent s confidentiality is protected. Respondent anonymity may be a key to response rates. Much literature on survey research is available. A literature or Web search or book on survey research will help to develop and guide the survey research skills of the investigator. Survey Introduction The introductory information is presented at the beginning (obviously) of the survey. The survey introduction to a subject (respondent) is accomplished in two ways. The introduction, of course, depends on which survey format is used. For paper or Internet surveys, the introductory information 95 is the first thing the subject should see. If a survey is sent by email with an or a Web URL, the introductory information can be in the body of the email. For telephone or face-to-face delivery, the introductory information is read or provided in some physical way such as an agreement to participate. (For telephone surveys, the subject should voice his or her agreement to participate.) Regardless of the manner of introduction, certain basic points must be included. A well-worded, informative introduction helps to increase the response rate. If the subject understands the purpose and need for their input or information, they are more likely to respond.1 The key points for an effective introduction are summarized in TABLES 7.4 and 7.5. A well- defined and informative introduction may induce subjects to participate by appealing to their sense TABLE 7.4 Key Points for the Introductory Statement or Cover Letter Key Point Example Who you are Describe yourself and/or organization and/or business in the opening paragraph Background for study In our practice, we have been using an appointment system . . . Purpose of the study We want to learn . . . What will be done with the data The information you provide will help us to . . . Why they were chosen We are asking you to take part because . . . Assurance of confidentiality You do not have to identify yourself in any way. All information will be reported in aggregate. All responses will be coded for study purposes. What respondents need to do Please take a few minutes to complete the survey. Please follow the instructions for each question. The person respondents can contact if they have questions If you have any questions or concerns, please contact . . . Approvals for the survey, if applicable This study has been approved by the Institutional Review Board of . . . (continues) 96 Section 2 The Research Process: Design TABLE 7.4 Key Points for the Introductory Statement or Cover Letter (continued) Key Point Example Return deadlines If possible, please complete and submit the survey by . . . Statement of importance The information you provide will be helpful in determining . . . Acknowledge support This study is supported by a grant from . . . This study is being conducted for ABC, Inc. Statement of appreciation We greatly appreciate your help in this matter and for taking the time to . . . Closing Sincerely, Your name, degree, titles Signature Co-investigators, degree, title of responsibility and the importance of their information and input. The introduction s tatement should be concise, well written, informative, and easy to understand, as exemplified in EXHIBIT 7.1. The level of wording, grammar, and writing style TABLE 7.5 Construction of a Hard Copy Cover Letter Construction Point Explanation Limit to one page. The longer the letter is, the less likely it is to be read completely. Use common block format with a readable font (e.g., Times New Roman, font size 12). Script, serif, calligraphy, and the like are hard to read. White, beige, or light gray paper, black print Easy to read; visually pleasing. Common language. Readability and ease of understanding are key. should fit the subjects to be surveyed. Unless you are dealing with highly educated professional people, the introduction statement (like newspapers) EXHIBIT 7.1 Example of an Introductory Statement from the Author s Research Dear , I invite you to participate in a survey on the importance of ranking health professions programs. You may access this survey by clicking on the URL below. You were selected to participate in this investigation by random selection from your professional organization s membership list. You do not have to identify yourself in any manner. Results will be reported in aggregate form. This is the first investigation on the opinion of health professionals on the ranking of educational programs. Your input will guide us in recommendations to the public and organizations. This investigation was approved as exempt by the IRB of State University. Please contact me at the address below if you have any questions. Thank you, J. D. Blessing Chapter 7 Survey Research should target an audience with middle school reading ability. Proofreading and editing are essential. It is a good idea to have individuals who are not part of the investigation read and interpret the introduction. An investigator may choose individuals who are representative of the subject sample or population to review the introduction and provide feedback. For either choice, include the introductory statement in any pre-study evaluations or pretests of the survey instrument and pilot studies. Follow-Up Follow-up notices or reminders to complete and return surveys may help increase response rates. Use a follow-up email, letter, or postcard. Be aware of reminders for surveys that deal with sensitive or very private information. For example, an investigator is studying the rate of sexually transmitted diseases (STDs) in corporate executives who travel frequently. Any reminder must be private and protected from outside eyes. Good wording for such a reminder is This is to remind you to take a few minutes to complete and return the survey we recently sent you. In the follow-up, add or use language that will encourage a response, such as, Your opinion is very valuable to us, and we want to include your input in our study. The wording of reminders should be general and neutral to encourage survey completion and return. Do not threaten or even appear to be negative. For example, do not say, If you do not complete and return the survey, no one will care what you think or know about STDs. A follow-up or reminder should have enough information for the recipient to identify the survey. Alternately, the survey (or link to the survey) can be re-sent with the survey reminder note. How many reminders to send and when to send them should be part of the decision making that is done in the initial methodology planning for the research project. Although there are no fixed rules for follow-ups, 10 to 14 days is usually reasonable. Subsequent reminders can be sent at four weeks. More than two reminders are probably a waste of time. 97 EXHIBIT 7.2 Example of a Reminder via Email or Mail REMINDER A few weeks ago, you received a Learning Assessment Survey from me. It was distributed through your program via email. If you have not completed the survey, please take a few minutes to do so and hit the submit button. If you need the survey site, please contact me. The information you provide through this survey will be helpful to me in my study of how students learn and, hopefully, to education in the future. Thank you for your time. Dennis Blessing, State University Phone number Email address Follow-up reminders can be done in several ways. One way is to send a reminder to everyone who was sent a survey. The other way is more economical if surveys have identification codes. If the survey is coded, the investigator can identify who has returned a survey and send follow-up reminders only to those subjects who have not responded. An explanation must be done for any codes that appear on a survey or a reminder. Subjects should be informed of what the number is and how it is used. Remember, if some type of code is used, the respondents are not anonymous. It is very important that their information be kept confidential. EXHIBIT 7.2 is an example of a reminder message used by the author. Types of Surveys Web-Based Surveys Web-based surveys have become the most common delivery method used to gather data on healthcare professionals. Typically, the survey is distributed using an email that contains a Web address (URL) for the survey or through a thirdparty survey platform. Two key advantages to 98 Section 2 The Research Process: Design Web-based surveys are: (1) the sample size can be large and dispersed, and (2) a large amount of data can be collected. Another advantage of Web-based surveys is that results can be downloaded to a database, eliminating data entry and its associated errors, time, and costs. Major limitations of Webbased surveys are the potential for a low response rate and respondent misinterpretation of items. In addition, the target sample must have a computer with online access. TABLE 7.6 lists some of the positives and negatives of Web-based surveys. An obvious advantage of Web-based surveys is that they can incorporate visually attractive TABLE 7.6 Aspects of Web-Based Surveys Aspect Comment Cost Low; usual cost is a fee required by the Web company Personnel Investigators only; requires no training of interviewers or field workers photographs and illustrations in color that are difficult and costly to reproduce in print. A Webbased survey can be longer without increasing the costs of delivery. In fact, the cost of administering a Web-based survey is the hosting fee, which is not dependent on the amount of activity at the site. A number of commercial company s design surveys and help with data collection. Typically, these sites are user-friendly and economical if the investigator does not have the time, skill, or tools to develop their own surveys .
You can get our ideal online homework help services with four easy steps. All you have to do is inform us of your requirements. Immediately we will match with a specialist in your academic discipline. After that, we will track your order, review it to ensure it meets your requirements, and submit it to your email.
Fill in your assignment instructions on the order form
Immediately you contact us for homework help; we will assign your assignment to an academic expert.
We have a team of editors, who will thoroughly review your assignment, ensuring there are no errors before submission
Once we have completed the assignment, we will deliver it to your account. Additionally, we will send a notification to your email.
Once you consult us for homework help online services, you will attest to why we are among the USA's best essay writing services.
Read on and find out some of the features that make our services superlative.
You can confirm that the content you get is original for free using our plagiarism checker.
We offer unlimited access to our essay homework samples.
Once you place an order, we format it accordingly without demanding additional charges.
We offer free unlimited revisions until your assignment is of your desired quality..
We have a team of specialized editors who review your assignment before submission.
You can track the progress of your homework for free.