Theme 4: Quantitative research
Paper
analysis:
The quantitative paper
i chose to read was “Social Networking Sites: Their Users and Social
Implications — A
Longitudinal Study ” by Petter Bae Brandtzæg. The study compares people that
use social networking sites (SNS) to people that do not for the purpose of
seeing the differences in social capital, here defined as amount of loneliness
felt, numbers of acquaintances, Face-to-face interactions and Bridging capital.
The quantitative study was conducted with the help of a survey site called
Norstat that according to the study seems reliable with a good mix of different
minded people in a wide range of ages. The survey was made three times over a
period of three years where the same people participated in all the surveys.
2000 people responded to the first wave of surveys, and the amount of people
that used SNS and responded to all surveys was 391. In an attempt to ensure
their data was reliable they made an investigation to see if the dropouts were
all from a specific gender, age or had a specific educational background. Their
conclusion was that the dropouts did not make the investigation that much less
reliable. In addition to that they studied already available material in order
to for a hypothesis and to see if what kind of differences were made compared
to their study.
I would argue that
the benefits of using a website for surveys online are that it is really cheap compared
to other options and easy to perform the surveys. As the site was a website
where the users were registered on the site for the sole purpose of making a
contribution to studies I think that the responses get more thought put into
and the correspondent takes the studies more seriously. However I would think
that making a study online still makes for less well thought off answers,
shorter answers and more people making the survey without being serious about
it than if the study was performed offline.
Some of the
problems I see with this study is obviously the dropout rate. I think that
categorizing dropouts only after gender, age and education does not ensure that
some subgroup were not more likely to make dropout (which would corrupt the
data). Another problem might be that some subgroup of people does not want to
respond to the survey at all. Some subgroups might be overrepresented in the
study while the goal of the study seems to be to represent the whole population
in Norway. This is obviously very hard problems to solve, you need to be more
active in making sure those who took the study the first year follow ups the
study the next year (perhaps by offering them more points on the site). The
second problem is even harder to solve, first of all you do not know what
subgroups might respond differently, what subgroups will not make the survey or
if there are any subgroups that might make flaws in the data at all. I think
they made a good job considering all in the study.
Which
are the benefits and limitations of using quantitative methods?
Quantitative methods are very useful when performing
research with many correspondents for the purpose of investigating or proving
some hypothesis by taking in a large number of correspondents. The quantitative
method requires a well-formed hypothesis and research questions in order to
produce a useful result. Therefore you have to have a good insight of the
problem or scenario you are investigating before investigating it further.
Limitations are that a quantitative study (for example online study) can never
be performed with the same quality as a qualitative study because it would be
to costly and correspondent would lose interest in responding to a to much time
consuming study.
Which
are the benefits and limitations of using qualitative methods?
Qualitative methods are on the other hand useful when
you are investigating a complex or previously uninvestigated phenomena. This
method is useful in producing hypothesis and gain insight on some problem. The
method allows to produce a deeper investigation with more complex questions,
longer interviews and perhaps more follow-ups and research material (provide
correspondents with material required to perform the research). However the
qualitative method is to a larger extend a subjective study where you have to
rely on a few correspondents and perhaps different twists in the analysis
depending on the people performing the interviews as well as the few
correspondents picked.
Hi Martin,
SvaraRaderaI find it very interesting that only 391 out of 2000 people that responded used a SNS. I agree with you that it might not be enough to categorize the dropouts after age, gender and education. I think that the geographical factor and their income are relevant to look at as well. I also think that a online survey might not be suitable to represent a whole population. But like you say, these are very tricky problems and probably hard to solve.
Keep up the good work!
Sofia
Hello Sofia and thank you for the comment.
RaderaTo clarify a little, 391 out of 2000 responded to all the 3 survey given out yearly over 3 years. Some of those who got the study did not respond at all, some did only respond to 1-2 surveys. I very much think that geografical factor and income would be suitable things to consider when investigating the participants (the location was norway in this case). Though i think they made a pretty good selection of people for the study, if you read the study you can see that they were using participants from a wide range of ages and education (somewhat correlated to income)
Actually I use to participate in different studies by Norstat, they have a Swedish site called panel.se were you gain points for every survey you complete. The points can be exchanged to gift checks for example. Is it written anything about if the participants in this study gained any points or other compensation for completed surveys? If that is the case, doesn’t this influence what types of people who participate in the study? I feel for example that I probably not will fill in any more surveys when I finally get a full time job and don’t have need for gift checks anymore…
SvaraRaderaI believe the only thing was written was some argumentation why Norstat was a reliable representation of the population (which i bought at the time) though i believe it was told in the lecture or seminar that participants get compensation (something like 30 SEK per hour). It is an interesting thought that those who are well paid is less likely to respond, the study does not really take that into account. Someone can be well educated but not well paid (i.e. a senior student), possible flaw there.
SvaraRaderaThough the study takes different people from different ages and different educational background it is uncertain wether a large group of the "well educated" correspondents consists of senior students which might make the illusion of a well represented population but in reality excludes well educated seniors.