To better understand
this week theme about quantitative methods we had a Lecture and a Lab exercise.
Lecture: Choosing Quantitative Methods with
Martha Cleveland
The Lecture
was very intense regarding the theme, full of examples that made me think once
again on how careful we need to be when designing a research, especially
considering unique or mixed methods and sampling.
Martha divided us in small groups for discussions
about advantages and disadvantages of mixed methods. As main positive aspects I
remember that the more data you have the chances to get a better and more
accurate result. Also that by setting up an experiment that can be replicable
increases the quality of the result once you can get either the same result,
validating previous ones or new results, showing opportunities to be further analyzed.
As a negative aspect I remember that framing the main research question and the
hypothesis might be hard considering different methods, and also that the
results might be hard to be connected one to each other.
Moreover
she said that the mixed research methods are being more and more used by researchers
and indicated a journal that is based only with this method.
Sampling
was also discussed in the Lecture, especially how hard it is to get a random
sample (I´ll try to look for a good definition and post on the blog later). It
was also very interesting to listen to her telling about how gender, age and intercultural
aspects should be taken in consideration while sampling.
LAB exercise, quantitative methods
Before we
went to the LAB a Lecture about how to perform an analysis of a research was
given, so we would understand its complexity and also how careful you need to
be to choose the right filters. We also were introduced to the SPSS Statistics
system that to me looked a lot like Excel. I have never seen this system before
and all the previous researches I have performed we analyzed the data in Excel.
The second
part of the Lab we went to the computers rooms to use the program. Although I
must confess that it was a little frustrating because the system did not work
properly due to some errors. Still we
were able to do one exercise in which we filtered a multiple answer question.
It was very nice to see how easily a report of the desired data could be
visualized in tables and exported.
To close
this week theme I thought of a very funny video that shows an experiment with
kids called the marshmallow experiment. Could there be any different outcome placing
a kid in front of a marshmallow?! Seems like the reactions vary a lot! Enjoy!
In quantitative research it is certainly good to have quite a large dataset. But, I would say that it is even more important to have high quality data. It is also possible to use some quantitative methods on quite small data sets.
ReplyDeleteWe are very sorry about the technical difficulties during the lab. I hope that you can find some other time to experiment more with the software. In some ways SPSS is similar to Excel, but it is much more powerful when doing statistical analyses. If you would choose to e.g. do a quantitative master theses, I recommend using SPSS. It is quite simple to use, when considering how powerful the software actually is.
Fernanda, I really like this video, and another one, with grown-ups in the same situation :)
ReplyDeleteActually I looked into various types of Sampling Methods already, I think that this site has the best definitions of all the types, Random especially, and gives the chance to understand the differences -> http://www.statpac.com/surveys/sampling.htm
Wikipedia is also very good, but also confusing (e.g. "simple random sampling should not be confused with random sampling" -> http://en.wikipedia.org/wiki/Simple_random_sample). Probably we should edit it :)
I would also like to add on your disadvantages of mixed research that it’s pretty hard to handle the financial issue. I can imagine that it’s pretty tough for researchers to find donators for mixed research. I also think it’s a huge challenge to find the right quantitative and qualitative tools in a mixed research which actually fit to the question and the topic. Also the amount of information has to be handled very well so you won’t draw wrong conclusions out of them.
ReplyDelete