Problems With a Telephone Survey

¶ … modal time for dropouts in live surveys either short or long is the half-point mark in the survey, the researcher needs to get to the most crucial information needed early in the survey (Cape, 2010). Choosing crucial information depends on the focus of the researcher (except for demographic information, see below). Moreover, the questionnaire is a bit cumbersome and will require quite a bit of training before interviewers can use it effectively (especially sections like section F). Perhaps it would be better to put in more guidelines, simplify some content, and make it a bit more user friendly. The fewer experimenter errors one makes the lower the bias and the better the reliability (and hence validity) of the measure (Bradburn, Sudman, & Wansink, 2004; Kalton & Schuman, 1982). Some of the issues with specific sections are discussed next.

Question A4 asks about reading within the past week, whereas other questions in the section ask about more typical behavior patterns over time. For purposes of validity it is best to keep questions measuring similar parameters consistent with regards to how often they are performed (Bradburn et al., 2004). Moreover, the optimal recall length participants have for ingrained behaviors such as reading the paper or watching the news is longer and more reliable than it is for other behaviors, so questions of this nature can be more general (Clarke, Fiebig, & Gerdtham, 2008).

In section B. The researcher uses a dichotomous opinion variable which could yield more information at a higher level of measurement (Bradburn et al., 2004). This can be problematic for later analysis with other variables that appear at the ordinal or interval level. While classic findings suggest no significant changes the internal consistency of a scale may result from this type of change (e.g., Matell & Jacoby, 1971), these changes do yield more information and it is best to remain as consistent as possible (Bradburn et al., 2004). If the variable will be used as a dichotomous independent variable it is better to measure it on an ordinal/interval scale for consistency and dichotomize later for analysis (Bradburn et al., 2004).

Section D. asks if the respondent or anyone in the household belongs to a particular group. If the researcher wants to know who actually belongs to the group in question this cannot be ascertained. Perhaps specifying who belongs to the group would be useful for later analysis (Bradburn et al., 2004).

Section E. asks a number of double-barreled questions. If the researcher wants to know more specific details regarding the particular behavior in question it would be prudent to add a query after each question to get more specific information (Bradburn et al., 2004).

Section L. has a number of confusing questions and is quite cumbersome (e.g., L2A). These really need to be simplified. Moreover, again questions use dichotomous answers that measure dimensional attitudes. For purposes of comparison and analysis later these variables could be transformed to ordinal/interval level (e.g., L4 how strongly do you favor & #8230; with a Likert scale response). This would also streamline the questionnaire and the interviewer would not have to jump around but only ask one question regarding if the participant opposes or favors some policy.

Section P. might be problematic in a live interview format because you may be putting some people on edge by making them feel ignorant. This can make them a bit defensive and as it is written the researcher does not even have demographic information on the participants yet. If this information is crucial to the aims of the research you might want to reword the questions to make them more opinion type questions as opposed to grilling them (e.g., who do you believe has the final responsibility & #8230; if they do not know then ask them if they care to guess).