AACR
AutoReport produces machine scoring of student
responses and generates instructor feedback reports. Upon uploading new student
data is scored using a
ensemble machine learning algorithm, the results of the scoring is analyzed and
the scoring, statistical analysis, and graphical representations are assembled
to interactive reports.
The
AACR Question Setup question set-up web app handles the input of new questions
into the AACR machine scoring system. This app trains scoring models, runs a
cross-validation analysis of the models, and records the necessary question
metadata needed to construct instructor feedback reports.
The
Exploratory Analysis app allows for the analysis of unscored student responses
using a number of analytic approaches. The resulting analysis may be used
during the early stages of rubric development to identify content and
structures in student data.
Confirmatory
Analysis allows for the analysis of hand-scored student responses using machine
learning tools. This app is useful for rubric refinement, the testing of
hand-scored data sets, and confirmation of predictive models prior to the
insertion of a question into the AACR machine scoring system.
The
path map app is a tool for the production of path map representations of
student progression through a sequence of questions.
This app may be used to strip HTML tags from Excel spreadsheets.
Data Uploading Tools
For the AACR research group to use with datasheets.