Data Set Information:

The data can be used to try to predict student learning in SE teamwork based on observation of their team activity

**** README FILE from the submitted data ZIP ****

#  San Francisco State University
#  Software Engineering Team Assessment and Prediction (SETAP) Project
#  Machine Learning Training Data File Version 0.7
#  ====================================================================
#
#  Copyright 2000-2017 by San Francisco State University, Dragutin
#  Petkovic, and Marc Sosnick-Perez.
#
#  CONTACT
#  -------
#  Professor Dragutin Petkovic:  petkovic '@' sfsu.edu
#
#  LICENSE
#  -------
#  This data is released under the Creative Commons Attribution-
#  NonCommercial 4.0 International license.  For more information,
#  please see
#  [Web link].
#
#  The research that has made this data possible has been funded in
#  part by NSF grant NSF-TUES1140172.
#
#  YOUR FEEDBACK IS WELCOME
#  ------------------------
#  We are interested in how this data is being used.  If you use it in
#  a research project, we would like to know how you are using the
#  data.  Please contact us at petkovic '@' sfsu.edu.
#
#
#  FILES INCLUDED IN DISTRIBUTION PACKAGE
#  ======================================
#  This archive contains the data collected by the SETAP Project.
#
#
#  More data about the SETAP project, data collection, and description
#  and use of machine learning to analyze the data can be found in the
#  following paper:
#
#  D. Petkovic, M. Sosnick-Perez, K. Okada, R. Todtenhoefer, S. Huang,
#  N. Miglani, A. Vigil: 'Using the Random Forest Classifier to Assess
#  and Predict Student Learning of Software Engineering Teamwork'.
#  Frontiers in Education FIE 2016, Erie, PA, 2016
#
#
#
#  See DATA DEscriptION below for more information about the data.  The
#  README file (which you are reading) contains project information
#  such as data collection techniques, data organization and field
#  naming convention.  In addition to the README file, the archive
#  contains a number of .csv files.  Each of these CSV files contains
#  data aggregated by team from the project (see below), paired with
#  that team's outcome for either the process or product component of
#  the team's evaluation.  The files are named using the following
#  convention:
#
#                  setap[Process|Product]T[1-11].csv
#
#  For example, the file setapProcessT5.csv contains the data for all
#  teams for time interval 5, paired with the outcome data for the
#  Process component of the team's evaluation.
#
#  Detailed information about the exact format of the .csv file may be
#  found in the csv files themselves.
#
#
#  DATA DEscriptION
#  ====================================================================
#  The following is a detailed description of the data contained in the
#  accompanying files.
#
#  INTRODUCTION
#  ------------
#
#  The data contained in these files were collected over a period of
#  several semesters from students engaged in software engineering
#  classes at San Francisco State University (class sections of CSC
#  640, CSC 648 and CSC 848).  All students consented to this data
#  being shared for research purposes provided no uniquely identifiable
#  information was contained in the distributed files.  The information
#  was collected through various means, with emphasis being placed on
#  the collection of objective, quantifiable information.  For more
#  information on the data collection procedures, please see the paper
#  referenced above.
#
#
#  PRIVACY
#  -------
#  The data contained in this file does not contain any information
#  which may be individually traced to a particular student who
#  participated in the study.
#
#
#  BRIEF DEscriptION OF DATA SOURCES AND DERIVATIONS
#  -------------------------------------------------
#  SAMs (Student Activity Measure) are collected for each student team
#  member during their participation in a software engineering class.
#  Student teams work together on a final class project, and comprise
#  5-6 students.  Teams that are made up of students from only one
#  school are labeled local teams.  Teams made up of students from more
#  than one school are labeled global teams.  SAMs are collected from:
#  weekly timecards, instructor observations, and software engineering
#  tool usage logs.  SAMs are then aggregated by team and time interval
#  (see next section) into TAMs (Team Activity Measure).  Outcomes are
#  determined at the end of the semester through evaluation of student
#  team work in two categories:  software engineering process (how well
#  the team applied best software engineering practices), and software
#  engineering product (the quality of the finished product the team
#  produced).  Thus for each team, two outcomes are determined, process
#  and product, respectively.  Outcomes are classified into two class
#  grades, A or F.  A represents teams that are at or above
#  expectations, F represents teams that are below expectations or need
#  attention.  For more information, please see the paper referenced
#  above.
#
#  The SE process and SE product outcomes represent ML training classes
#  and are to be considered separately, e.g. one should train ML for SE
#  process separately from training for SE product.
#
#  TIME INTERVALS FOR WHICH DATA IS COLLECTED
#  ------------------------------------------
#  Data collected continuously throughout the semester are aggregated
#  into different time intervals for the semester's project reflecting
#  different dynamics of teamwork during the class.  Time intervals
#  represent time periods in which a milestone was developed by each
#  team.  A milestone represents a major deliverable point in the class
#  for all student teams.  The milestones are roughly divided into the
#  following topics:
#
#            M1 - high level requirements and specs
#            M2 - more detailed requirements and specs
#            M3 - first prototype
#            M4 - beta release
#            M5 - final delivery
#
#  Time intervals are combinations of the time in which milestones are
#  being produced.  Time intervals are used in research only.
#
#  In addition to time intervals corresponding to milestones, a number
#  of time intervals combining multiple T1-T5 time intervals have been
#  calculated.  This was done to group student activities into design
#  vs. implementation phases which have different dynamics.
#
#  These time intervals are defined as follows:
#
#      Time Interval        Corresponding Milestone Periods in Class
#    -----------------    --------------------------------------------
#           0               Milestone 0
#           1               Milestone 1
#           2               Milestone 2
#           3               Milestone 3
#           4               Milestone 4
#           5               Milestone 5
#           6               Milestone 1 - Milestone 2 inclusive
#           7               Milestone 1 - Milestone 3 inclusive
#           8               Milestone 1 - Milestone 4 inclusive
#           9               Milestone 1 - Milestone 5 inclusive
#          10               Milestone 4 - Milestone 5 inclusive
#          11               Milestone 3 - Milestone 5 inclusive
#
#
#
#  SETAP PROJECT OVERALL DATA STATISTICS
#  ==================================================================
#  The following is a set of statistics about the entire dataset which
#  may be useful in the configuration of machine learning methods.
#
#  This data was collected only from students at SFSU.  Global teams
#  represent only the data from the SFSU student portion of the team.
#
#  GENERAL STATISTICS
#  ------------------
#                       Number of semesters: 7
#                            First semester: Fall 2012
#                             Last semester: Fall 2015
#                        Number of students: 383
#                            Class sections: 18
#
#                    Number of TAM features: 115
#         Number of class labels (outcomes): 2
#
#                     Issues closed on time:   202
#                        Issues closed late: +  53
#                                            -------
#                              Total issues:   255
#
#  TEAM COMPOSITION STATISTCS
#  --------------------------
#      Local Teams:    59
#     Global Teams: +  15
#                   ------
#            Total:    74 Teams
#
#  OUTCOME (CLASSIFICATION) STATISTICS
#  -----------------------------------
#   Total Outcomes: 74
#
#                Proces               Product
#           ------------------  ------------------
#  outcome:      A       F           A       F
#                49      25          42     32
#
#  TAM FEATURE NAMING CONVENTION
#  -----------------------------
#  A systematic approach to aggregating and naming TAM features was
#  developed.  By using this systematic approach, TAM feature names are
#  produced that are human understandable and intuitive and related to
#  aggregation method.
#
#
#  There are a number of base TAM which are then aggregated into
#  aggregated TAM.
#
#  base TAM
#  --------
#
#  General TAM
#  -----------
#  The following TAMs are collected for each team: Year, semester,
#  timeInterval, teamNumber, semesterId, teamMemberCount,
#  femaleTeamMembersPercent, teamLeadGender, teamDistribution
#
#  Calculated TAM
#  --------------
#  For each team, TAM were calculated from SAMs for every time interval
#  Ti.  The core TAM variables where for each we compute as applicable:
#  count, average, standard deviation over weeks, over students etc.
#
#  TAMs collected by Weekly Time Cards (WTS) TAM
#  ---------------------------------------------
#  teamMemberResponseCount, meetingHours, inPersonMeetingHours.
#  nonCodingDeliverablesHours, codingDeliverablesHours, helpHours,
#  globalLeadAdminHours, LeadAdminHoursResponseCount,
#  GlobalLeadAdminHoursResponseCount
#
#  TAMs collected  by Tool Logs (TL) TAM
#  -------------------------------------
#  commitCount, uniqueCommitMessageCount, uniqueCommitMessagePercent,
#  CommitMessageLength
#
#  Collected by Instructor Observations (IO) TAMs
#  ------------------------------------------------
#  issueCount, onTimeIssueCount, lateIssueCount
#
#
#  AGGREGATED TAM
#  --------------
#
#  Several aggregation method and derived variable names for TAMs
#  reflect how the core TAM variables were aggregated in final TAM
#  measures for each time interval Ti:
#
#  Let VAR be the core TAM variable above. The naming conventions and
#  aggregation operators to obtain TAMs for each time interval Ti were
#  as follows:
#
#Total - total sum of VAR in the time interval Ti
#Average - average of VAR in the time interval
#StandardDeviation - SD of variable in time interval
#Count - count of events measured by VAR (e.g. missed
#  checkpoints) in time interval
#  AverageByWeek - total sum/count of VAR in the time interval
#  divided by weeks in time interval
#  StandradDeviationByWeek - the standard devation of the weekly
#  total of VAR taken over the time interval
#  AverageByStudent - total count/sum of VAR in time interval,
#  divided by number of students in the team
#  StandardDeviationByStudent - standard deviation of  VAR in the
#  time interval, over students in the team
#
#
#  NULL VALUES
#  -----------
#  NULL values are used in the training data to indicate that no SAMs
#  were recorded in that particular time period, week, or for that
#  student.
#
#  Frequently TAM features involving teamLeadHours or globalTeamLead
#  hours will result in a NULL for a particular training sample.  For
#  local team leads, that usually means that the local team lead did
#  not complete any timecard surveys for the aggregation in quesiton.
#  While for global team lead TAM features this may also be the case,
#  the more usual cause of NULLS in global team lead TAM features comes
#  from the fact that most teams are not global, and therefore this
#  statistic was not gathered for these teams.
#
#  It is left to the individual researcher to decide how to accomodate
#  NULL values, and the data is included in this file.  Though these
#  may not be useful for machine learning directly, valuable
#  information can be obatined with some processing.
#
#  TAM FEATURES
#  ------------
#  The following is a list of tam features available in the data files.
#  The TAM feature names are listed in the order in which the data
#  appear in each training sample, i.e. the first feature corresponds
#  to the first column, the second feature corresponds to the second
#  column, etc.
#
#  The first sample line in the data section of the data file is not a
#  true sample, but consists of TAM feature names, which allows for
#  easy import into spreadsheets and for human readability.
#
#  The final two TAM features (columns) are the outcome data for
#  process and product, and are the last two columns in each sample
#  row.  The training sample data follow the header comment section.
#
#
#  TAM FEATURE LIST
#  ----------------
#  year
#  semester
#  timeInterval
#  teamNumber
#  semesterId
#  teamMemberCount
#  femaleTeamMembersPercent
#  teamLeadGender
#  teamDistribution
#  teamMemberResponseCount
#  meetingHoursTotal
#  meetingHoursAverage
#  meetingHoursStandardDeviation
#  inPersonMeetingHoursTotal
#  inPersonMeetingHoursAverage
#  inPersonMeetingHoursStandardDeviation
#  nonCodingDeliverablesHoursTotal
#  nonCodingDeliverablesHoursAverage
#  nonCodingDeliverablesHoursStandardDeviation
#  codingDeliverablesHoursTotal
#  codingDeliverablesHoursAverage
#  codingDeliverablesHoursStandardDeviation
#  helpHoursTotal
#  helpHoursAverage
#  helpHoursStandardDeviation
#  leadAdminHoursResponseCount
#  leadAdminHoursTotal
#  leadAdminHoursAverage
#  leadAdminHoursStandardDeviation
#  globalLeadAdminHoursResponseCount
#  globalLeadAdminHoursTotal
#  globalLeadAdminHoursAverage
#  globalLeadAdminHoursStandardDeviation
#  averageResponsesByWeek
#  standardDeviationResponsesByWeek
#  averageMeetingHoursTotalByWeek
#  standardDeviationMeetingHoursTotalByWeek
#  averageMeetingHoursAverageByWeek
#  standardDeviationMeetingHoursAverageByWeek
#  averageInPersonMeetingHoursTotalByWeek
#  standardDeviationInPersonMeetingHoursTotalByWeek
#  averageInPersonMeetingHoursAverageByWeek
#  standardDeviationInPersonMeetingHoursAverageByWeek
#  averageNonCodingDeliverablesHoursTotalByWeek
#  standardDeviationNonCodingDeliverablesHoursTotalByWeek
#  averageNonCodingDeliverablesHoursAverageByWeek
#  standardDeviationNonCodingDeliverablesHoursAverageByWeek
#  averageCodingDeliverablesHoursTotalByWeek
#  standardDeviationCodingDeliverablesHoursTotalByWeek
#  averageCodingDeliverablesHoursAverageByWeek
#  standardDeviationCodingDeliverablesHoursAverageByWeek
#  averageHelpHoursTotalByWeek
#  standardDeviationHelpHoursTotalByWeek
#  averageHelpHoursAverageByWeek
#  standardDeviationHelpHoursAverageByWeek
#  averageLeadAdminHoursResponseCountByWeek
#  standardDeviationLeadAdminHoursResponseCountByWeek
#  averageLeadAdminHoursTotalByWeek
#  standardDeviationLeadAdminHoursTotalByWeek
#  averageGlobalLeadAdminHoursResponseCountByWeek
#  standardDeviationGlobalLeadAdminHoursResponseCountByWeek
#  averageGlobalLeadAdminHoursTotalByWeek
#  standardDeviationGlobalLeadAdminHoursTotalByWeek
#  averageGlobalLeadAdminHoursAverageByWeek
#  standardDeviationGlobalLeadAdminHoursAverageByWeek
#  averageResponsesByStudent
#  standardDeviationResponsesByStudent
#  averageMeetingHoursTotalByStudent
#  standardDeviationMeetingHoursTotalByStudent
#  averageMeetingHoursAverageByStudent
#  standardDeviationMeetingHoursAverageByStudent
#  averageInPersonMeetingHoursTotalByStudent
#  standardDeviationInPersonMeetingHoursTotalByStudent
#  averageInPersonMeetingHoursAverageByStudent
#  standardDeviationInPersonMeetingHoursAverageByStudent
#  averageNonCodingDeliverablesHoursTotalByStudent
#  standardDeviationNonCodingDeliverablesHoursTotalByStudent
#  averageNonCodingDeliverablesHoursAverageByStudent
#  standardDeviationNonCodingDeliverablesHoursAverageByStudent
#  averageCodingDeliverablesHoursTotalByStudent
#  standardDeviationCodingDeliverablesHoursTotalByStudent
#  averageCodingDeliverablesHoursAverageByStudent
#  standardDeviationCodingDeliverablesHoursAverageByStudent
#  averageHelpHoursTotalByStudent
#  standardDeviationHelpHoursTotalByStudent
#  averageHelpHoursAverageByStudent
#  standardDeviationHelpHoursAverageByStudent
#  commitCount
#  uniqueCommitMessageCount
#  uniqueCommitMessagePercent
#  commitMessageLengthTotal
#  commitMessageLengthAverage
#  commitMessageLengthStandardDeviation
#  averageCommitCountByWeek
#  standardDeviationCommitCountByWeek
#  averageUniqueCommitMessageCountByWeek
#  standardDeviationUniqueCommitMessageCountByWeek
#  averageUniqueCommitMessagePercentByWeek
#  standardDeviationUniqueCommitMessagePercentByWeek
#  averageCommitMessageLengthTotalByWeek
#  standardDeviationCommitMessageLengthTotalByWeek
#  averageCommitCountByStudent
#  standardDeviationCommitCountByStudent
#  averageUniqueCommitMessageCountByStudent
#  standardDeviationUniqueCommitMessageCountByStudent
#  averageUniqueCommitMessagePercentByStudent
#  standardDeviationUniqueCommitMessagePercentByStudent
#  averageCommitMessageLengthTotalByStudent
#  standardDeviationCommitMessageLengthTotalByStudent
#  averageCommitMessageLengthAverageByStudent
#  standardDeviationCommitMessageLengthAverageByStudent
#  averageCommitMessageLengthStandardDeviationByStudent
#  issueCount
#  onTimeIssueCount
#  lateIssueCount
#  processLetterGrade
#  productLetterGrade

Attribute Information:

See above

Relevant Papers:

D. Petkovic, M. Sosnick-Pérez, K. Okada, R. Todtenhoefer, S. Huang, N. Miglani, A. Vigil: “Using the Random Forest Classifier to Assess and Predict Student Learning of Software Engineering Teamwork” Frontiers in Education FIE 2016, Erie, PA, 2016

Citation Request:

Please cite above FIE paper