Page 68 - Special Topic Session (STS) - Volume 3

STS515 Steve MacFeely
Teaching data science - a user perspective
or
(preparation for a career as an official
statistician)
Steve MacFeely 1,2
1 United Nations Conference on Trade and Development, Geneva, Switzerland
2 Centre for Policy Studies, University College Cork, Ireland

Abstract
Within the broader panoply of statistics and data science, official statistics is a
discipline of its own. While it requires the same core skills as most other
statistical professions, it also requires other dimensions or elements unique to
official statistics. Most formal statistical and data scientist training has to date,
not addressed these other elements, and therefore have not equipped
students very well for the complex and challenging world of official statistics.
Introducing students to this fascinating world in university, would allow them
to better understand how statistics can play an integral role in public policy
and the important role they could play in shaping the way we think about our
world. This paper discusses, from the perspective of an employer of data
scientist and statistical graduates, some of the topics that universities might
consider including in their curricula. Furthermore, the argument that a greater
focus on nurturing competencies rather than on developing specific skills to
prepare students for a rapidly changing world is explored.

Keywords
official statistics; statistical literacy; skills; competencies

1. Introduction

Q: What’s the difference between a data scientist and a statistician?
A: 40K per annum

For the purposes of this paper, I am treating a data scientist and a statistician
as being synonymous (I won’t be addressing salary differentials). Perhaps if we
dive deep enough there are some subtle differences, but for all practical
purposes I believe they are essentially the same profession requiring the same
skills and competencies. The real difference perhaps is that one – the data
scientist has jumped on the ‘data revolution’ wave and has rebaptised him or
herself for marketing reasons. For the remainder of the paper, I will refer to
statisticians, but the reader can substitute data scientist if they prefer.
It might seem obvious, but a statistician hoping to work as an official
statistician in a National Statistical Office (NSO) or National Statistical System
(NSS) should be statistically literate or should at least be capable of becoming
57 | I S I W S C 2 0 1 9