dropR
DropR analyzes data from Internet-based experiments for differences in dropout (aka attrition, mortality, break-off) between conditions. Currently, DropR supports visual inspection of dropout, Odds ratio by item, Chi Square tests of differences at any particular item (e.g. test for overall dropout difference), Kaplan-Maier survival estimation, Rho family tests.
DropR follows a simple step-by-step process. Begin by uploading your data under 'Upload your data' or choose our demo file and choose 'experimental_condition' as the grouping variable under 'Identify' to see what DropR can do for you.
Plot options
Hints
- color blind and printer friendly palettes support up to 8 different categories (colors).
- When re-labelling conditions, use comma (,) as a seperator. Make sure to list as many names as conditions selected.
- With dropR you can produce graphs for publication in various formats. You may choose from vector formats such as .svg and .pdf and the .png format for rendered pixels. While size is relevant to any format, resolution only applies to .png and will be ignored when vector formats are chosen.
Upload your data
1 Choose
a .csv file from your disk
- Indicate whether the first line of your data is meant to be a header.
- Choose the proper column delimiter, text quote and missing value coding for your file. Note that, proper missing values (empty fields) are accepted in addition by default.
- Make sure to use reasonable coding for missing values (i.e. empty cells). Avoid ambigous coding such as -99, -999 etc.
- Check the preview window below. Iff your data is displayed as expected you are good to start with your analysis. DON'T forget to hit 'update data' in the right box when you're ready.
or use a demo dataset instead
2 Specify
.csv properties
3 Identify
questions and conditions
Data Preview
χ2-test options
Test outcomes
Odds ratio by item
Kaplan-Meier Estimation
Kaplan-Meier survival curve
Test Survival Curve Differences
About
DropR is a joint project by Ulf-Dietrich Reips and Matthias Bannert that followed naturally from the long-standing need in Internet science and online research for methods and tools to address the fact that dropout (aka attrition, mortality, break-off) occurs much more frequently when conducted via the Internet than in traditional lab-based research.