Back when I was doing research in psychology, one of my biggest challenges has been enough data. And while these rates are probably still do for many experimental researchers with the ability of the Internet to access data, it appears that the data faster than rabbits reproduce.
Now I see a lot of research on the comparison with the administration and analysis of vast amounts of data to be overcome.
Even a moderate number of variables can lead to endless variations on the analysis. "Hmm, I wonder if thestem length is correlated with the wind direction. What is the measure of plant size, leaf area? And 'this correlates with the direction of the wind? "This can go on forever.
According to Frank Scarpaci, owner of Design Works Project, there is a
"1:10:100 rule:
Every dollar of planning and preparation will save $ 10 or $ 100 is project work to solve problems after the project. "
And as in science, time is money (or even more valuable) Planning StatisticsThe analysis saves time and endless frustration. I mean, you think better for an hour now the analysis as a "once in a year after the auditors rip to shreds, right?
The best time for planning the analysis is to collect data as before. This prevents the (all too common) situations where one has understood, requires a different variable, or if you have something measured on a different scale. Grant applications force you to do this, but would benefit in each study.
How do you wantReally? They are based on the results, we report. You should already know the questions to be answered in this study, but written with it in a list will keep you on track. It was not necessary (and should) respond to any questions that the data in this study.
I found a great place for a simple analysis of the plan is a brilliant article by Darryl Bem newspaper articles written about writing. The article is very good (and I highly recommend), but the most usefulPlanning is the section "Presentation of the results." This section outlines seven steps for reporting the findings. For planning, I condense these three:
1. State of the conceptual assumptions Ask
2. To formulate this hypothesis in terms of variables that measure the concept
3. List of statistical test or method to answer this question
Just repeat together to answer these three phases for all the hypotheses of the study. Start with the general andimportant, and work down from there.
Do these three steps before sitting down to analyze three advantages:
1. It forces you, the variables, use the selection. The election at the beginning you can bring unwanted variables from large data sets, time in a much faster processing. It defines the variables to make the univariate analysis, and clearly defined which variables to collect.
2. It is frightening is irrelevant analysis of implementation, which saves time, energy,and frustration as well as your article clearly and logically.
3. It makes the results in writing a children's section.
It 'true that there is a place for exploratory data analysis and some surprises, there are still pop-up data, a trend unexpected, and so on. But again and again on the track is always easier when the direction in which you are going to know.
For more information, see:
Bem, D. (2003). "Writing the Empirical Journal Article". In The CompleatAcademic: a practical guide for the beginning social scientist, 2 Edition. Darley, JM, Zanna, MP and Roediger III, HL (Eds.), Washington, DC: American Psychological Association.
No comments:
Post a Comment