1 of 6

{ Udacity × Bertelsmann } Data Science Scholarship Project Showcase

Challenge Course

A Roadmap of Data Science Beginnings

Summer 2018 Expedition

Melissa

Simon

2 of 6

Is every lesson of the Challenge Course* equally dense?

2

HYPOTHESIS

Python and SQL lessons contain more content than they appear relative to the overall course.

{

}

QUESTION

ASSUMPTIONS

  • Progression through course at consistent pace.
  • No significant prior subject matter knowledge assumed.
  • Scoring of difficulty relative to only activities within this course.
  • Scoring of activity duration assumes learner has a satisfactory understanding of the applied concepts.

* Bertelsmann Data Science Challenge Scholarship Course (2018)

3 of 6

Data collected from the Course was analyzed in spreadsheets and Tableau.

3

1496

DATA ROWS

OBJECTIVE MEASURES

video length

word count

​

SUBJECTIVE MEASURES

activity duration

activity difficulty level

AGGREGATION

various levels

sections

lessons

units*

time

CALCULATED

len(video)�+

COUNT(word)

+

{activity:duration}

* 4 units — meta: 1, 22, 33; statistics: 2-21; Python: 23-27; SQL: 28-32

METHOD

4 of 6

4

Programming units are denser and each unit requires about ⅓ of the time.

GIVEN COURSE COMPOSITION (%)

ANALYZED COURSE COMPOSITION (%)

ANALYZED TIME PER CONTENT TYPE PER LESSON

4 units — meta: 1, 22, 33; statistics: 2-21; Python: 23-27; SQL: 28-32

FINDING

5 of 6

Allocate more time for Python & SQL and do not rely on progress indicator.

5

difficulty — level 1: low; level 2: moderate; level 3: high�based on experience and perception of Forum & Slack learner questions

RECOMMENDATIONS

TIME PER SECTION�grouped by units

  • INSTRUCTORS�adjust course progress indicator (%).�
  • COURSE PARTICIPANTS allocate more time for Python and SQL lessons.

MEAN DIFFICULTY�grouped by lessons

stats

m

Python

SQL

m

m

6 of 6

{ powered by }

Bertelsmann Data Science Challenge Scholarship Course

July 2018

6

{ Udacity × Bertelsmann } Data Science Scholarship Project Showcase