Another "data management & research database for dummies" question
November 27, 2014 12:42 AM Subscribe
I'm undertaking a research project in which I'm evaluating several different organizations based on specific yet qualitative criteria, e.g. "promotes outreach", "reports on progress", etc. I need to able to capture the information and then display the results in some sort of meaningful way so that it can be incorporated into a report.
Excel would seem the obvious answer here, but because of the heavily text-based input rather than quantitative (numeric) data I feel like a spreadsheet is not really the way to go. I've looked into various database options, but there are so many that I feel overwhelmed and not sure where to start. Oh, and I'd prefer a Windows-based solution. I'm not sure what other information I should provide in order to get recommendations, so please feel free to ask! Also please assume I'm completely ignorant about data management solutions, if that wasn't already glaringly, embarrassingly apparent.
Excel would seem the obvious answer here, but because of the heavily text-based input rather than quantitative (numeric) data I feel like a spreadsheet is not really the way to go. I've looked into various database options, but there are so many that I feel overwhelmed and not sure where to start. Oh, and I'd prefer a Windows-based solution. I'm not sure what other information I should provide in order to get recommendations, so please feel free to ask! Also please assume I'm completely ignorant about data management solutions, if that wasn't already glaringly, embarrassingly apparent.
Note that a database, for all intents and purposes, is a way to collect lots of spreadsheets together. If you want to categorise and display your data you will need to quantify it in some way eventually. For that reason, and your professed ignorance of data management, I would probably recommend something like excel. I don't know how large your data set is going to be, which will determine to some extent what tools you should be using.
Even though you might be using text as your response you might want to think about putting it into categories. So reading your notes and saying that this bit essentially means that college A: promotes outreach.
I can imagine various options for this. Imagine the following spreadsheet:
College|outreach methods used|reports on progress|others
A|google ads|monthly reports|...
A|Television ads|accounting on request|...
A|Radio ads|0|..
A|leaflets|0|..
B|...
Where a 0 here indicates missing: i.e. these are all the methods you recorded. Alternatively you could make one row per college:
College|outreach:googleads|outreach:Radioads|outreachanyother|reports:yearly/monthly/NA|...
A|1|1|1|Monthly|..
B|1|0|1|Yearly|..
C|0|1|0|NA|....
Of these two, the former gives you more flexibility in that you can expand your list to allow for the college, while the latter will be more useful for doing some quantitive analysis (46% of colleges use google ads, while only 10% use radio ads). Depending on the quantity of your data, you could start with the first method then transition to the second once you got a good grasp of the number of categories that might exist.
posted by Cannon Fodder at 3:36 AM on November 27, 2014
Even though you might be using text as your response you might want to think about putting it into categories. So reading your notes and saying that this bit essentially means that college A: promotes outreach.
I can imagine various options for this. Imagine the following spreadsheet:
College|outreach methods used|reports on progress|others
A|google ads|monthly reports|...
A|Television ads|accounting on request|...
A|Radio ads|0|..
A|leaflets|0|..
B|...
Where a 0 here indicates missing: i.e. these are all the methods you recorded. Alternatively you could make one row per college:
College|outreach:googleads|outreach:Radioads|outreachanyother|reports:yearly/monthly/NA|...
A|1|1|1|Monthly|..
B|1|0|1|Yearly|..
C|0|1|0|NA|....
Of these two, the former gives you more flexibility in that you can expand your list to allow for the college, while the latter will be more useful for doing some quantitive analysis (46% of colleges use google ads, while only 10% use radio ads). Depending on the quantity of your data, you could start with the first method then transition to the second once you got a good grasp of the number of categories that might exist.
posted by Cannon Fodder at 3:36 AM on November 27, 2014
Are you quantifying these text examples? E.g., an org is rated 8.5 for transparency and then you list the supporting quotes?
posted by michaelh at 6:33 AM on November 27, 2014
posted by michaelh at 6:33 AM on November 27, 2014
Response by poster: At the moment, nope, I hadn't planned to assign numeric values, just possibly identify trends or commonalities.
Thanks all for the answers so far, this is helpful.
posted by ladybird at 12:24 PM on November 27, 2014
Thanks all for the answers so far, this is helpful.
posted by ladybird at 12:24 PM on November 27, 2014
NVIVO is a horrible piece of software (bloated, slow, a bit tough to learn). I would consider alternatives as well.
Back when I was doing research for money I taught myself perl to help with organisation, consolidation and analysis of rich data sets. It certainly opened up some interesting career opportinities for me.
posted by singingfish at 1:57 PM on November 27, 2014
Back when I was doing research for money I taught myself perl to help with organisation, consolidation and analysis of rich data sets. It certainly opened up some interesting career opportinities for me.
posted by singingfish at 1:57 PM on November 27, 2014
« Older TV, movie, and book suggestions for a Daria fan. | Love-Hate Letters to a genre / trope /idea? Newer »
This thread is closed to new comments.
If you're looking for a Windows recommendation the best I can suggest is really going back to basics. Write all the information you have about the organisations in a word document. From there highlight information that is connected in different colours. So anything that provides evidence that an organisation 'promotes outreach' is highlighted blue, while 'reports on progress' is highlighted green.
This information that you're highlighting should be direct evidence that you've collected eg: In October 2014 organisation x released a 20 page document through their website which reported on their progress in meeting the expectations of the board. Note that this one sentence can be highlighted in a number of ways, it provides evidence of reporting on progress, that the org has a board and that it reports via the web. So it would have a number of highlights.
When you're finished, pull out all the information highlighted in the same colours into separate documents/sections, making sure the info is tagged with information about the organisation. This way when you're looking at all the information together you can see similarities between organisations that are doing the same thing.
Eg - the majority of the organisations that promoted outreach were (based in the inner city/focused on youth/received government funding) while few of the organisations that promoted outreach were (based in rural areas/received donations).
posted by bernardbeta at 2:33 AM on November 27, 2014 [1 favorite]