Sometimes details on outcome variable is not provided how to deal with such data.
What you might do depends on the type of effect size you are using, what is missing, et cetera. Most effect sizes can be calculated several different ways. Provide details and we'll try to help.
Dear Smita
I faced the same problem when I conducted a meta-analysis.
I suggest the following:
1. Contact the primary investigator or the corresponding author by email and ask them to provide the missing data.
2. There are ways to calculate some of the missing data. Please have a look at the Cochrane handbook for systematic reviews (http://handbook-5-1.cochrane.org/front_page.htm)
The following are some useful links from the Cochrane handout:
7.7.3 Data extraction for continuous outcomes
http://handbook-5-1.cochrane.org/chapter_7/7_7_3_data_extraction_for_continuous_outcomes.htm
7.7.3.2 Obtaining standard deviations from standard errors and confidence intervals for group means
http://handbook-5-1.cochrane.org/chapter_7/7_7_3_2_obtaining_standard_deviations_from_standard_errors_and.htm
7.7.3.3 Obtaining standard deviations from standard errors, confidence intervals, t values and P values for differences in means
http://handbook-5-1.cochrane.org/chapter_7/7_7_3_3_obtaining_standard_deviations_from_standard_errors.htm
7.7.7 Data extraction for estimates of effects
http://handbook-5-1.cochrane.org/chapter_7/7_7_7_data_extraction_for_estimates_of_effects.htm
7.7.7.2 Obtaining standard errors from confidence intervals and P values: absolute (difference) measures
http://handbook-5-1.cochrane.org/chapter_7/7_7_7_2_obtaining_standard_errors_from_confidence_intervals_and.htm
7.7.7.3 Obtaining standard errors from confidence intervals and P values: ratio measures
http://handbook-5-1.cochrane.org/chapter_7/7_7_7_3_obtaining_standard_errors_from_confidence_intervals_and.htm
I hope this will help.
Mansour
Hi Smita,
Gordon is correct: the strategy largely depends on the type of effect size you are using and what data are missing (is is the effect size itself or the variance)?
There are lots of ways to account for missing data, each with its pros and cons. With some more information about your situation, we can give you some more specific help!
-Halle
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis