
I need a statistician to analyze my data according to journal editors comment
5154
$$
- Posted:
- Proposals: 4
- Remote
- #145622
- Archived
Description
Experience Level: Intermediate
Hi I have recently submitted a paper for publication but i got some comment from the editor and need the help of a statistician. this is the comment that we need to deal with .
There is one overriding problem with the data treatment that needs to be addressed, probably with the help of a statistician. The disattenuation formula seems to be misapplied in this study and appears to have resulted in an overcorrection of the correlations. Perfect correlations (r=1.0) should always be suspect, particularly in an educational study where additional variables are hard to control. Further examination of the data in this report shows this to be the case. For example, using the data from Table 2 and Table 3A, the rc for "MCQ Understanding" v. "FR Understanding" actually calculates to 1.0897, a so-call impossible correlation. This clearly shows that the "True correlations" in Table 3B are overcorrected. One good guideline is that Spearman's formula (the one used in this study) should not be used, without the risk of overcorrection, if reliability estimates drop below 0.7 (Reference: J.W. Osborne. http://pareonline.net/getvn.asp?v=8&n=11) . Three of the six Cronbach's alpha reliability estimates in Table 2 are indeed below 0.7 and are yielding corrected correlation coefficients that are too good to be true. The Osborne reference suggests other means of dealing with this situation. The conclusions reached by the authors using just the uncorrected r's or properly corrected r's may still be the same, only not as strongly, but more realistically, supported by the numbers.
There is one overriding problem with the data treatment that needs to be addressed, probably with the help of a statistician. The disattenuation formula seems to be misapplied in this study and appears to have resulted in an overcorrection of the correlations. Perfect correlations (r=1.0) should always be suspect, particularly in an educational study where additional variables are hard to control. Further examination of the data in this report shows this to be the case. For example, using the data from Table 2 and Table 3A, the rc for "MCQ Understanding" v. "FR Understanding" actually calculates to 1.0897, a so-call impossible correlation. This clearly shows that the "True correlations" in Table 3B are overcorrected. One good guideline is that Spearman's formula (the one used in this study) should not be used, without the risk of overcorrection, if reliability estimates drop below 0.7 (Reference: J.W. Osborne. http://pareonline.net/getvn.asp?v=8&n=11) . Three of the six Cronbach's alpha reliability estimates in Table 2 are indeed below 0.7 and are yielding corrected correlation coefficients that are too good to be true. The Osborne reference suggests other means of dealing with this situation. The conclusions reached by the authors using just the uncorrected r's or properly corrected r's may still be the same, only not as strongly, but more realistically, supported by the numbers.
Hassan S.
0% (0)Projects Completed
5
Freelancers worked with
5
Projects awarded
57%
Last project
9 Aug 2012
Saudi Arabia
New Proposal
Login to your account and send a proposal now to get this project.
Log inClarification Board Ask a Question
-
There are no clarification messages.
We collect cookies to enable the proper functioning and security of our website, and to enhance your experience. By clicking on 'Accept All Cookies', you consent to the use of these cookies. You can change your 'Cookies Settings' at any time. For more information, please read ourCookie Policy
Cookie Settings
Accept All Cookies