Three Reasons Education And Social Scientists Prefer Proprietary Software And Data Formats | Kyle D. Husmann, Ph.D

"When I first jumped into the world of education and social science research, I was surprised by the extent to which proprietary software and data formats were used to manage and exchange research data. As someone passionate about the use of open software and open data in scientific research, this was troubling to me.

Everywhere I looked I saw data being shipped around in SPSS, SAS, and Stata formats. What gives? Proprietary software makes analyses less repeatable, the opposite of what you want for Open Science. Proprietary data formats make data less accessible, less interoperable, and harder to reuse… the complete opposite of FAIR data principles. (Although to be fair, Stata’s dta format is an exception on this front, because at least they openly document their format’s internals).

At first, I thought the landscape was like this because of the nice graphical interfaces provided by proprietary statistical software (as opposed to coding in R or Python), as well as the network effects of vendor lock-in. And don’t get me wrong, these are big reasons! But after over a half decade of building education and social science data management and analysis pipelines using exclusively open software, I have come to appreciate other details that aren’t so obvious at first glance...."