A few of the Stata datasets used for testing are not in the format implied by their file names: - [stata4_114.dta](https://github.com/pandas-dev/pandas/blob/ab2980ba52ae24db72765b3c4ae1d33ac8fad3f5/pandas/tests/io/data/stata/stata4_114.dta) appears to actually be saved in 115 format. - [stata9_115.dta](https://github.com/pandas-dev/pandas/blob/ab2980ba52ae24db72765b3c4ae1d33ac8fad3f5/pandas/tests/io/data/stata/stata9_115.dta) appears to actually be saved in 117 format - [stata10_115.dta](https://github.com/pandas-dev/pandas/blob/ab2980ba52ae24db72765b3c4ae1d33ac8fad3f5/pandas/tests/io/data/stata/stata10_115.dta) appears to actually be saved in 117 format This may result in reduced test coverage for the expected versions.