Research: Chapter 13: Survey Designs

Common Uses of Cross-Sectional Survey Designs

A survey: 1.of current attitudes, beliefs, opinions or practices. 2.to compare two or more groups. 3.To measure community needs 4.A statewide/national survey

Approaches to Survey Data Collection

Recall: two basic formats employed for collecting survey data: •The Questionnaire •The Interview

Steps for Conducting Survey Research

Step 1: Decide what you want to learn. Step 2: Determine your population and sample selection procedures. Step 3: Determine the survey methodology and data collections procedures you will employ. Step 4: Develop a questionnaire or locate one used in previous research. Step 5: Pilot test the questionnaire. Step 6: Administer the questionnaire. Step 7: Analyze the data. Step 8: Write the report.

Types of Survey Design

Two basic design types: •Longitudinal: occurs when data is collected over time. •Cross-sectional: data is collected at one point in time.

Disadvantages

While an interpersonal bond may be established with the use of one on one interviews, some participants will be uncomfortable with particular topics and the lack of anonymity may interfere with securing honest responses.

Survey Research

Within the counseling field, surveys are commonly used by researchers both for the relative ease in which such studies can be conceptualized and for the directness in which data related to some aspect of human life can be obtained.

One-On-One Interviews

•A researcher or interviewer asks a predetermined set of closed ended questions and records the responses. •A well trained interviewer will be able to: -Be polite and build rapport with participants. -Provide instructions and ask questions as they are written. -Complete the interview in the allotted time frame while avoiding inserting personal opinions and maintaining confidentiality.

Telephone Interviews

•Allows for collecting data from participants widely scattered among different locations. •More difficult to establish a bond with the participant over the phone. •The nonverbal behaviors of the participants cannot be observed.

Mailed Questionnaires

•An example of questionnaire that is mailed to a sample and typically includes a letter of explanation and a return envelope. •Relatively inexpensive and allows the participant to complete it in his own time.

Maximizing Survey Response Rates

•Be certain that individuals targeted as participants are knowledgeable about the topic. •Pre-notify participants they will soon be receiving a survey. •Provide a cover letter, either written or electronic with the survey. •Decide upon and systematically implement follow up procedures. •Use incentives if possible to encourage participation.

Longitudinal Survey Design

•Data collected is more detailed, richer, and provides greater depth. •Three Basic Varieties: 1.Trend Study: used to understand how a variable changes over time with a general population. Surveys members each year in the same population. The disadvantage is lack of information regarding the changes in the attitudes of specific individuals. 2.Cohort Studies: surveys representatives of a group over time to follow the group's progress. Has the advantage off examining and understanding an entire group but reveals nothing about specific individuals or insight into the variations of responses. 3.Panel Studies: Most rigorous of approaches that surveys the same individuals each time of data collection. Provides focused information about specific participants over an extended time. Cost may be prohibitive and it is often difficult to locate specific persons over time.

Interview Surveys

•High degree of interaction between the researcher and the participant allowing for a connection and the potential of increased depth of information. •Encourages the participant to complete the survey. •Allows an opportunity to ask questions or make comments.

Purpose of Survey Research

•Important method of measurement in social science research. •Knowledgeable participants report on personal experiences of interest to the researcher. •Data is effective for exploring trends within populations, describe relationships between variables or compare groups. Survey studies may fall into one of the following categories: 1.To describe an existing phenomenon by gathering detailed information about it. 2.To identify peoples problems or concerns. 3.To give good reason for and support current circumstances or practices. 4.To make comparisons or evaluation. 5.To determine what others are doing with similar problems or situations.

Designing Survey Items

•Items for a survey questionnaire should be accurate, unbiased, and easy for respondents to answer. •It is critical to tailor items that will result in meaningful data collection. •First consideration is determining the most appropriate question structure between open-ended and closed-ended questions.

Cross-Sectional Survey Designs

•Most commonly used design for survey research. •Data is collected at only one point in time. •Data is collected from a sample then results are utilized for a variety of purposes. •Data is collected quickly and provides information about current day attitudes.

Disadvantages of Electronic Questionnaires

•No meaningful connection between the researcher and the participants. •Multiple requests and follow-ups are to be expected.

Challenges of Mailed Questionnaires

•Participants may not return surveys as there is no personal connection between the participant and the research. •No opportunity for participants to ask questions so researchers should attempt to confirm: -Are the directions unmistakably clear? -Are the items well-worded and the meaning unambiguous? -Are the scales easy to complete and the layout user friendly? -How will you conduct follow up?

Electronic Questionnaires

•Popular due to the potential to reach a large number of participants. •May be established with a web based source or email format. •Benefits include: -Reduced cost -Ease and speed of administration -Anonymity of participants -Access to larger samples -Ability to target minority populations

Self Administered Questionnaires

•Regardless of design, all questionnaires have a form for the participant to complete and return to the researcher by hand, mail, or electronically. •Asks a set of predetermined questions to be answered by hand or on the computer.

Questionnaires and Interviews

•Regardless of purpose, data is generally collected by questionnaires or interviews. •Questionnaires are self administered by the participants using pencil and paper or computer. •Interviews are administered verbally by a trained interviewer.

Descriptive Research

•Survey research falls under the general category of descriptive research. •The goal is to describe or explain a participant's opinions or preferences related to a phenomenon. •Survey methods gather information about a group by asking individuals for their feedback or reviewing documentation related to them.

Limitations of Survey Research

•Surveys are not useful for determining cause of a phenomena. •Results depend on participants truthfully and accurately reporting attitudes, opinions and behaviors. •Errors of omission, confusion, or false memory are impossible to avoid as are bias. •Survey data cannot substitute for expert judgments and analysis. •Require careful preparation, are labor intensive, and can take much time to implement & analyze.

Closed-Ended Questions

•The respondent is asked a question and provided a selection of answers. -Ex: Which of the following factors most contributed to your decision to attend the University of Georgia? ▪Proximity to where I live ▪Reputation of the University ▪Family members who attended the university. ▪The cost of tuition

Open-Ended Questions

•The respondent is asked a question that provides an opportunity to give an expansive answer. -Ex: What led you to attend the University of Georgia? ▪May result in inadequate responses due to no interviewer available to encourage in-depth disclosure.

Response Rates

•The total percentage of questionnaires returned called response rate. •A high return rate is always the goal. •The higher the response rate, the greater chance that the results actually reflect attitudes, opinions, and experiences of the population. •Differences between the people who responded and those who choose not to is called non-response bias. •A response rate of 25% is considered low. •A response rate of 75% or greater is considered high. •Response rates of 50% or higher are often found in published survey research. •To maximize response rate verify that individuals selected have knowledge of the area under investigation and are reasonably motivated.

Guidelines for Optimizing Effectiveness of Survey Items

•Use simple words and complete sentences. •Use as few words as possible and make questions technically accurate. •Avoid items with multiple questions and soften questions with risky content. •Provide an option of "not applicable" but avoid "check all that apply". •Avoid overlapping answers.

Question Construction

•Writing clear and concise questions is difficult. •The length of the question, sophistication of the language used, educational level of respondent, length of the survey, style of the question and available options for responding must all be considered.