SmartGPA: How Smartphones Can Assess and Predict Academic - - PowerPoint PPT Presentation

SmartGPA: How Smartphones Can Assess and Predict Academic Performance of College Students

Rui Wang, Peilin Hao, Xia Zhou, Andrew Campbell (Dartmouth College) Gabriella Harari (University of Texas at Austin)

performance

behaviors features correlations prediction

level change

studentlife.cs.dartmouth.edu studentlife.cs.dartmouth.edu

we extend studentlife

behaviors features correlations prediction

class attendance, studying and partying

semantics of location

studying

studying

study areas

sound activity

focus

study duration (hours)

6 12 18 24

attendance

0.25 0.5 0.75 1

week

1 2 3 4 5 6 7 8 9

attendance study

midterm

attending classes and studying

partying

partying

party places

sound

activity co-location

party duration (hours)

1.5 3 4.5 6

week

1 2 3 4 5 6 7 8 9 10

partying trends across the term

mid term green key

party duration

0.4 0.8

weekday

1 2 3 4 5 6 7

study duration

1.7 3.4

weekday

1 2 3 4 5 6 7

behaviors features correlations prediction

capturing the dynamics of behavior

how to represent the data?

study duration (hour) 1.25 2.5 3.75 5 week 1 2 3 4 5 6 7 8 9

study duration (hour) 1.25 2.5 3.75 5 week 1 2 3 4 5 6 7 8 9

use mean to measure level

mean = 3.15

behavior term slope

study duration (hour) 1.25 2.5 3.75 5 week 1 2 3 4 5 6 7 8 9

term slope = 0.29

behavior term slope

study duration (hour) 1.25 2.5 3.75 5 week 1 2 3 4 5 6 7 8 9

term slope = 0.29

midterm

study duration 2 6 10 14 week 1 2 3 4 5 6 7 8 9

pre/post midterm slope

pre-slope = 2.23 post-slope = -0.86

midterm

study duration (scaled) 0.35 0.7 week 1 2 3 4 5 6 7 8 9

breakpoint — when students change their behavior to adapt

student 1 student 2 different breakpoint

• iteratively select every

week as breakpoint

• use one or two linear

regressions to fit the data before and after the breakpoint study duration 2 4 week 1 2 3 4 5 6 7 8 9

breakpoint — how to compute

• iteratively select every

week as breakpoint

• use one or two linear

regressions to fit the data before and after the breakpoint study duration 2 4 week 1 2 3 4 5 6 7 8 9

breakpoint — how to compute

study duration 2 4 week 1 2 3 4 5 6 7 8 9

• iteratively select every

week as breakpoint

• use one or two linear

regressions to fit the data before and after the breakpoint

breakpoint — how to compute

study duration 2 4 week 1 2 3 4 5 6 7 8 9

breakpoint — how to compute

week 1 2 3 4 5 6 7 8 9 week 1 2 3 4 5 6 7 8 9

we use Bayes Information Criterion to select the breakpoint MSE1 MSE5 MSE6

behaviors features correlations prediction

which of the 193 features relate to performance?

studying, partying and GPA

study duration study focus - activity study focus - audio party duration R value

• 0.45
• 0.3
• 0.15

0.15 0.3 0.45 0.6

studying, partying changes and GPA

pre-midterm class attendance pre-midterm study duration after-midterm conversation duration R value 0.113 0.225 0.338 0.45

behaviors features correlations prediction

what models can predict GPA?

studying partying activity conversation stress / positive affect mental health personality

+

GPA

…

w0 w1 w2 w3 wi+1 wi+2 wi+3

use lasso to regularize training

leave-one-out cross validation

selected features

three sensor-based behavioral features

• conversation duration night breakpoint
• conversation duration evening term-slope
• study duration

three EMA features

• positive affect
• positive affect post-slope
• stress term-slope
• ne personality
• conscientiousness

prediction performance

CDF

0.25 0.5 0.75 1

absolute error

0.1 0.2 0.3 0.4 0.5

MAE = 0.179

goodness of fit:

• R2 = 0.559
• r = 0.81, p < 0.01
• ur model can

distinguish high and lower performers

Thanks, I’m done