What Mode for Your Survey?

There are three modes of administering surveys, they are face-to-face, telephone, and self-administered (on paper or through the computer). How do you decide which to use? There is no single best method; all have advantages and disadvantages, which are summarized in the table below (modified from Vogt et al., 2012, p. 20).

                                                                                Mode of administration

Advantages and disadvantages	Face-to-Face	Telephone	Self-Administered
1. Usable with respondents who cannot read	Yes	Yes	No
2. Researcher can explain the meaning of a question	Yes	Yes	No
3. Researcher knows identity of respondent	Yes	Yes	No
4. Interaction with administer of survey	High	Medium	Low
5. Certainty that all respondents get exactly the same question	Low	Medium	High
6. Cost per respondent	High	Medium	Low
7. Time per question	High	High	Low
8. Possible to administer in groups	Yes	No	Yes
9. Effort required to obtain a large sample size	High	Medium	Low
10. Problems of access to respondents in remote or unsafe areas	Yes	No	No
11. Researcher's efficiency tied to that of postal, internet, or telephone system	No	Yes	Yes

As the researcher, you need to weigh each of these considerations, keeping both the advantages and disadvantages in mind.

Next time I will post an updated blog index. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

A Research Opportunity!

Drs. Amy Sickel and Lee Stadtlander, faculty at Walden University, are seeking doctoral level students and faculty as participants for a research study. This study will be validating a new survey on doctoral writing and writing efficacy (self-confidence). You will be asked to complete a series of 4 surveys related to research and doctoral writing. This 30-minute study is completely confidential and may give you some insights into your own writing and research skills. As a thank you for participating, the first 100 participants will receive a $10 Starbuck's gift card. If you are interested, please go to https://www.surveymonkey.com/r/validate_writing_survey

The study will close on March 17, 2017

Survey questions

Continuing our discussion of surveys:

4. You know how you will use the answers.

You need to have a clear idea as to how you will use the survey questions to address your research questions. Keep in mind that short surveys are generally much better than long ones in keeping participants focused and finishing the survey.

Each of your research questions typically relate to a number of survey questions. It is a good idea to match up your research and interview questions to make sure that you will actually be able to answer your research questions. If you have survey questions without a matching research question, you probably do not need the survey questions. If you have a research question without a survey question, you need to add more survey questions. Here is an example of matching a research question and survey questions. 

Research Questions	Survey Questions
1. How do patients' searches, for health related information on the internet, relate to visits to their primary health provider?	1. In the past month, have you searched for health related information on the internet? 2. Before your most recent visit to your primary health provider, did you search for any health related information on the internet? 3. After your most recent visit to your primary health provider, did you search for any health related information on the internet?

You should also have a plan as to how you will analyze the data you collect. Without such a plan, you may come to realize that you have collected hundreds of surveys with a missing variable that is needed for your analysis. As an example, if you would like to analyze your data using gender as an independent variable, make sure you have included it in the survey.

5. You can expect an adequate response rate.

An otherwise excellent survey study can be ruined by a low response rate. The difficulty is that if only a small proportion of the people you ask to be in the survey actually complete it, you can no longer generalize to the population of interest. For this reason, if you suspect a low response rate is likely, you would be better off having fewer people and then interview them.

What is a good response rate? There is not a clear answer to this. You will need to review the literature and see what others report. There are a few things you can do to improve your response rate. Using a university's participant pool, generally results in a high rate of response (keep in mind they are not representative of the general population). If you can guarantee respondents that their answers will be anonymous (the researcher does not know who responded), it generally helps. The second best choice is guaranteeing confidentiality (only the researcher knows who responded).

Next time we will continue our discussion of surveys. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

Survey Research

Surveys are popular, because they are efficient; researchers can gather a great deal of information for a small cost and generally in a short amount of time. To consider whether surveys are a good idea for your study, consider these issues (from Vogt, Gardner & Haeffele (2012).

1. The data are best obtained directly from the respondents

2. Your data can be obtained by brief answers to structured questions.

3. You can expect respondents to give you reliable information.

4. You know how you will use the answers.

5. You can expect an adequate response rate.

Now let us consider each of these issues.

1. The data are best obtained directly from the respondents.

Often the only way to obtain information about people is by asking them. This is especially true of subjective data, i.e., the inner states of the subjects being studied, such as attitudes, beliefs, or values. You can also use surveys to collect objective data, such as respondents' age, income, or years of work experience, which could be obtained in other ways, but it is usually easier simply to ask. However, if you have access to records, such as educational and medical records, these may be more accurate and possibly more efficient than asking individuals. People's memories are notoriously poor on such questions, as "how often have you visited your doctor in the past year." Having access to their records would make it much more accurate. Keep in mind that asking unimportant or needless questions tend to annoy participants and they may quit your study because of it.

2. Your data can be obtained by brief answers to structured questions.

Surveys work best when asking structured questions. Structure refers to the degree of control you want to exercise over the answers you will get. The questions may be either completely unstructured and vague or highly structured. The more structured the questions, the more participants are constrained to your predetermined answers. Therefore, you may be limiting their responses; they may have other things to say. On the other hand, having broad and exploratory questions as in interviews, may not be the best choice for surveys, as the data may not be able to be summarized.

3. You can expect respondents to give you reliable information.

Sometimes surveys are an unwise choice because respondents may find it difficult to answer accurately or truthfully. Perhaps the information you seek is hard to remember or too sensitive for people to be willing to reveal it to a stranger. It is often assumed that social desirability biases much survey research, meaning that respondents may answer questions to reflect what they think is socially appropriate rather than what they really believe or do. Racial beliefs and attitudes provide an example of this. An example given by Vogt et al. (2012) is in the second half of the 20th century, the number of people answering national surveys in ways indicating they held a racial prejudice dropped rapidly compared to earlier surveys. Whether people were less prejudiced in the 1990s than in the 1950s is uncertain. However, it is clear that the behavior of respondents on the same survey questions dramatically changed.

Next time we will continue our discussion of surveys. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

Surveys: Structure and Truth

Surveys work best when asking structured questions. Structure refers to the degree of control you want to exercise over the answers you will get. The questions may be either completely unstructured and vague or highly structured. The more structured the questions, the more participants are constrained to your predetermined answers. Therefore, you may be limiting their responses; they may have other things to say. On the other hand, having broad and exploratory questions as in interviews, may not be the best choice for surveys, as the data may unable to be summarized.

Sometimes surveys are an unwise choice because respondents may find it difficult to answer accurately or truthfully. Perhaps the information you seek is hard to remember or too sensitive for people to be willing to reveal it to a stranger. It is often assumed that social desirability biases much survey research, meaning that respondents may answer questions to reflect what they think is socially appropriate rather than what they really believe or do. Racial beliefs and attitudes provide an example of this. In the second half of the 20^th century, the number of people answering national surveys in ways indicating they held a racial prejudice dropped rapidly compared to earlier surveys. Whether people were less prejudiced in the 1990s than in the 1950s is uncertain. However, it is clear that the behavior of respondents on the same survey questions dramatically changed. 

Next time we will continue our discussion of surveys. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

Resisting Temptation

You have started your study, but you are not getting many participants. A small little voice in your head suggests that it would be so easy to make up the data… no one would know… When that voice starts whispering to you, hear my voice shouting, "DON'T DO IT!" There are so many ramifications of falsifying data, let us begin with a recent story in the news. Donald Green, a researcher at Columbia University and Michael LaCour, published a study  December 2015 in Science magazine. In the study, Green and LaCour reported results that suggested that voters’ support for same-sex marriage increases following a 20-minute conversation with a gay advocate for same-sex marriage.

Intriguing results are they not? Other researchers attempted to replicate the results. When they were unable to do so, they contacted the original researchers and the journal and wanted to know more about the original data. Green could not produce his data, as more questions were asked; it was found that he lied about funding sources also.

What are the implications of your making up data? You can have your Ph.D. revoked, be dismissed from the university (with no ability to return), if you received any grant funding you could face criminal charges for fraud. In addition, your professional career is finished before you started. You would never be able to use anyone from your institution as a reference, as a result getting a job or entering another program would be difficult to impossible.

Recently IRBs have begun randomly auditing data (particularly collected from students). Falsifying data is not worth it. As a professional, you have a responsibility to be ethical in all aspects of your study. If you are having problems getting enough participants, talk to your committee. If your data is not coming out as expected, talk to your committee. There are ethical solutions, let the experienced researchers guide you to them.

Next time we will consider more aspects of surveys. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

When to use surveys?

Surveys are popular, because they are efficient; researchers can gather a great deal of information for a small cost and generally in a short amount of time. To consider whether surveys are a good idea for your study consider these issues (from Vogt, Gardner & Haeffele (2012):

1. The data are best of obtained directly from the respondents
2. Your data can be obtained by brief answers to structured questions.
3. You can expect respondents to give you a reliable information.
4. You know how you will use the answers.
5. You can expect inadequate response rate.

Over the next few posts, I will be examining these criteria. Today we will address the first issue whether the data are best obtained directly from the respondents.

Often the only way to obtain information about people is by asking them. This is especially true of subjective data, i.e., the inner states of the subjects being studied, such as attitudes, beliefs, or values. You can also use surveys to collect objective data, such as respondents' age, income, or years of work experience, which could be obtained in other ways, but it is usually easier simply to ask. However, if you have access to records, such as educational and medical records, these may be more accurate and possibly more efficient than asking individuals. People's memories are notoriously poor on such questions, as "how often have you visited your doctor in the past year." Having access to their records would make it much more accurate. Keep in mind that asking unimportant or needless questions tends to annoy participants and they may quit your study because of it.

Next time we will consider resisting temptation. Do you have an issue or a question that you would like me to discuss in a future post? Would you like to be a guest writer? Send me your ideas! leann.stadtlander@waldenu.edu

Vogt, W. P., Gardner, D. C., & Haeffele, L. M. (2012). When to use what research design. NY: The Guilford Press.