Page 159 - Data Science class 11
P. 159

Many researchers are tempted to do much of their data collection online. But, the fact is that online surveys are not
            always the most appropriate method of data collection, especially if respondents are in remote areas. If important
            respondents are left out of the survey, the result would suffer greatly.

            Paper Surveys
            As the name suggests, this survey uses the conventional paper and pencil technique. Despite paper surveys being a
            thing of the past, they are quite handy when it comes to field research and data collection. These surveys are usually
            conducted in remote areas since, in these places, computers, laptops, or other handheld devices are not accessible.
            This type of survey is the most expensive method of data collection as it involves deploying a large pool of human
            resources along  with time and money.

            Telephonic Surveys
            Telephonic  surveys  are  surveys  that  are  conducted  by  researchers  by  telephone.  Respondents  answer  questions
            related to the research topic as chosen by the researcher. These surveys are time-consuming, inconvenient, error-
            prone, and sometimes non-conclusive. The success of such surveys depends on the number of people willing to
            answer the questions on the phone and those willing to invest their time in the same.

            Interviews
            This is one of the most popular types of surveys. Such surveys are also known as in-person surveys or household
            surveys. This is a favourite choice amongst researchers since they get a chance to have face-to-face interaction with
            individuals. This method of surveying may seem outdated since we have online surveying at our disposal. Nevertheless,
            interviews still serve a purpose.
            Researchers  organise  interviews  when  they  want  to  discuss  something  personal  or  confidential  with  people.  By
            putting interviewees at ease, the interviewers may get genuine responses. Researchers often have questions that may
            necessitate extensive scrutiny to unravel the truth.

            Focus Groups
            These types of surveys are conducted in-person. Rather than just one individual, focus groups take into account a
            number of people. This group is usually small yet demographically diverse and led by a moderator. Focus groups
            are useful for sampling new products or having an exchange of views on a particular topic, often a contentious one.
            The aim of a focus group survey is generally to assess people’s responses to a product in a group setting, or to get
            people debating—with the moderator taking notes. This survey is an expensive method as a trained moderator must
            be  paid.  Moreover,  these  can  only  be  conducted  in  specific  locations  depending  upon  the  demographics  of  the
            participants. The participants also have to be heavily incentivised in the form of gift cards or money to show up.

            Kiosk Surveys
            Kiosk surveys happen on a digital screen at a physical location. These may pop-up in eateries, shopping outlets,
            hotel lobbies, hospitals, office spaces, etc.—anywhere a researcher or marketer wants to collect data from customers
            or  passers-by.  Kiosk  surveys  are  quite  effective  and  potent  methods  of  conducting  surveys  since  they  provide
            feedback instantly following a purchase or an interaction. These surveys enable us to gather responses as soon as the
            respondent experiences something or makes an opinion about a product or service. As such, kiosk surveys ensure
            genuine responses.

            4.3.3 is the research a Survey, an Observational Study, or an experiment?

            In broad terms, a survey simply measures variables, an observational study attempts to find a relationship between
            variables, and an experiment attempts to establish a cause-and-effect relationship between Variables.






                                                                                              Randomisation    157
   154   155   156   157   158   159   160   161   162   163   164