Survey Data Collection Methods

telephone survey (person administered surveys)

This method is best when physical/face to face contact is not necessary. 1) Fast turnaround, quicker, you can do several surveys per hour (no travel time) 2) Quality control over sample. If random digit dialing (RDD) procedure is used, telephone survey may produce a representative sample. 3) Calls are way cheaper than face to face (no travel costs) 4) Long distance/geographically spread out respondents can be reached more easily. Disadvantages: No possibility of showing products or prototypes People are not good at remembering different response options and often just pick the last one (recency effect) (vs. primacy effect)

pros of hybrid surveys

multiple advantages to achieve data collection goal

person-administered surveys

"interviewer" reads questions to respondent and records answers either face-to-face, or over the telephone This was the primary mode of survey research for many years. However, as costs have increased and technology has advanced, it has lost most of its popularity, but they are still used when justified/needed

choice of survey method

- Time & money available for data collection (online fastest and cheapest) - Type of respondent interaction required (personal face to face needed for high interaction) - Incidence rate -Cultural/infrastructure considerations (no phone, don't want someone in home)

cons of hybrid surveys

1. mode affects response ex: would consumers answer differently in home vs. online 2. additional complexity: different types of surveys wording has to be adjusted more work for the researcher

survey

An interview with a large number of respondents using a predesigned questionnaire In recent decades, people are a lot less willing to participate in survey research Significant advances in technology and software programs make survey research a lot more efficient → but has not solved the nonresponse problem (e.g., rise of fun online quizzes (e.g., Buzzfeed): new wave of marketing research)

Cons of person administered surveys

Human error Slow speed High cost "Interview evaluation" (apprehensive they are answering the question "correctly". Feeling like they are being evaluated) Human error: A researcher could ask questions out of sequence, change the wording of a question, etc. especially when the researcher is bored or tired themself. Slow speed: human (person-administered) data collection is very slow, not as quick as online or phone High cost: often personal interviewers or administers/researchers are highly skilled and trained, therefore expensive (more expensive than doing online surveys) Risk of "interview evaluation": Interview evaluation means respondents can sometimes feel like they are being evaluated on their answers (even when that's not the case, and even when the administer/researcher is a total stranger). The person who is interviewed might be scared to make mistakes or say stupid things. Especially a problem with sensitive topics such as hygiene, finances, political opinions, etc.

Pros of person administered surveys

Immediate feedback Trust Quality control Adaptability Immediate feedback: If respondents do not understand a question or a concept, or they become distracted during the face to face (verbal) survey → the researcher/administer can adjust the question based on verbal or nonverbal feedback. When a respondent looks bored, the administer can say that "it won't take long anymore" to keep them motivated. Researcher can anticipate certain things based on respondents' body language etc. Trust: a good researcher/administer can create trust and understanding (which is something that a more "impersonal" form of data collection can not) Quality control: e.g. If someone responds incorrectly on some very obvious characteristic, the researcher/administer can see it right away and correct it. Obviously since the researcher/administer is there, he can watch over the process and make sure everything is correct, no questions are omitted, whether respondents are distracted, etc. → this way, more control over the quality of the data gathered. Adaptability: how to 'translate'/reword a question to a certain respondent, e.g., when working with kids, elderly people, foreign language speakers, etc., sometimes a little bit of adaptation could be needed.

mall survey (person administered surveys)

People are surveyed while shopping (walking in a mall) → problem of non-representative sample: only people who happen to be at the mall at that time (depending on what your research objective is, that might not be the typical target group you're going for). People can feel uncomfortable answering questions in an open public place (others staring), they carry bags of clothes and stuff, have places to go, are in a hurry, etc.

self-administered surveys

Pros: 1. reduced cost (no human administer to be paid 2. self-paced, own pace- more relaxed 3. no interview evaluation (no bias) Disadvantages: 1. respondent controls the survey (may not finish it, may do it way too late, may not hand it in) 2. lack of monitoring on researchers end, can't help respondent if they have a question/confused 3. high questionnaire requirements (all pressure on questionnaire to be completely self explanatory and clear)

computer-assisted surveys

Pros: 1. speed 2. less-error prone (computers do not become tired like humans) 3. use of pictures/videos (advertising pretesting) 4. real-time capture of data (immediately entered into database) 5. reduction of "interview evaluation" concern

in home survey (person administered surveys)

Researchers must travel from and to respondents' home → high cost To justify the high cost of the in-home interview, the personal contact must be essential to the success of the interview e.g., a company develops a new type of kitchen appliance → needs to be tested in home before survey questions can be asked e.g., respondents might not be able (due to illness, immobility, etc.) to travel somewhere to participate in research

in office survey (person administered surveys)

Takes place in the respondent's office or company lounge area, which is great for busy executives who are busy bossing around all day and have no time to spare (so you visit them). You can even bring a prototype/physical object related to your survey, e.g., if you want them to test something, try something, or even taste something (e.g., potential investor for your new hummus dip, or another business idea). Relatively high cost because travel costs are high if you have to visit respondents in their office. GATEKEEPERS Gatekeepers are people who are in between you (as the researcher/administer) and your respondent, for example a personal assistant or a front desk employee who may/may not give you further access to your respondent. They can give or deny you access.

standardization (advantages of surveys)

all respondents are asked the same questions in the same way/format, because the questionnaire is preset/predesigned, and organized in a rather fixed manner. Uniformity is crucial if you want to be able to compare answers across individuals.

ease of administration (advantages of surveys)

another person can easily administer the survey, or the respondent can fill out the survey alone, if left unattended (e.g., online survey: not much effort required from the researcher, you can just send it out online and wait til the results come in).

Mixed-mode (hybrid) survey

combination of 2 or more methods for data collection increasingly popular in the past years, because marketing researchers realize that respondents should be treated like customers: matching the data collection mode to the convenience of the respondent. Best of both worlds.

suitable for statistical analysis (advantages of surveys)

easy to analyze because it's so standardized, plus there's computer programs (like SPSS) that help with statistical analyses even with very large sample sizes.

tap into the unseen (advantages of surveys)

get beneath the surface, surveys can ask about respondents' motivations, reasons, etc. (although not to the same extent as focus groups or depth interviews)

sensitivity to group differences (advantages of surveys)

respondents can be divided into subgroups or segments based on certain characteristics → to compare across groups (meaningful differences)

mail survey (self-administered surveys)

survey mailed out and returned by mail 2 major problems: 1. nonresponse 2. self-selection bias those who *do* fill it out are probably the most interested/satisfied customers, or have the most time on their hands → not representative of the population! Due to this self-selection bias, you'll end up with a non-representative sample and thus, distorted data.

incidence rate

the percentage of potential respondents that qualify for a study based upon the given screening qualifications. Almost every survey or market research study involves some type of parameters or screener qualification questions. Incidence rate = the % of the population that possesses a certain characteristic that is necessary in order to be included in the survey Rarely research projects are targeted at "everyone" In most cases, there are qualifiers/prereqs in order to participate in a survey e.g. Registered voters, adults age 18 and over, etc. Sometimes incidence rate is very low: e.g. Only men above 60 with high cholesterol (above 250 level) low: spread out, not concentrated