“…“…But What Do the Top-Rated But What Do the Top-Rated Schools Do?”Schools Do?”
A Survey of A Survey of Introductory Introductory
Computer Computer Science Science CurriculaCurricula
www.cs.duke.edu/csed/openwiki/doku.php?id=teaching:startwww.cs.duke.edu/csed/openwiki/doku.php?id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE 2007
Covington, KY2007-03-09 @ 10:30
EST
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
Jeff ForbesJeff ForbesDuke UnivDuke Univ
Dan GarciaDan GarciaUC BerkeleyUC Berkeley
2/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
MotivationMotivation What do top PhD-What do top PhD-granting schools do granting schools do in their intro CS in their intro CS curricula?curricula? What courses What courses offered?offered?
Who teaches them?Who teaches them? What is taught?What is taught? How are they run?How are they run? Who are students?Who are students? What is the What is the innovation?innovation?
Who are “sep at Who are “sep at birth”?birth”?
We had questions!We had questions!
Computer Computer sciencescience
3/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
SurveySurvey We surveyed top 32 We surveyed top 32
schoolsschools 33 Qs in 7 categories:33 Qs in 7 categories:
InstitutionInstitution StaffStaff DemographicsDemographics ContentContent DeliveryDelivery StyleStyle MetaMeta
MethodMethod Swept web sites for infoSwept web sites for info Emailed text questions Emailed text questions
to faculty at each to faculty at each schoolschool
Built SurveyMonkey formBuilt SurveyMonkey form Phone calls Phone calls Email follow-upsEmail follow-ups
4/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Schools (2006 “US News & Schools (2006 “US News & WR”)WR”)
1.1. CMUCMUMITMITStanfordStanfordUC BerkeleyUC Berkeley
5.5. CornellCornellUI Urbana-ChampaignUI Urbana-Champaign
7.7. U WashingtonU Washington8.8. PrincetonPrinceton9.9. UT AustinUT Austin
U Wisconsin-MadisonU Wisconsin-Madison11.11. Cal TechCal Tech
Georgia TechGeorgia Tech13.13. UC San DiegoUC San Diego
U Maryland - College U Maryland - College ParkPark
15.15. HarvardHarvardUCLAUCLA
15.15. U Michigan-Ann ArborU Michigan-Ann Arbor18.18. ColumbiaColumbia
Purdue - West Purdue - West LafayetteLafayetteU PennU PennYaleYale
22.22. BrownBrownRiceRiceUNC - Chapel HillUNC - Chapel Hill
25.25. DukeDukeU Mass - AmherstU Mass - AmherstUSCUSC
28.28. Johns HopkinsJohns Hopkins29.29. NYUNYU
Rutgers - New Rutgers - New BrunswickBrunswickUC IrvineUC IrvineVirginiaVirginia
5/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
DisclaimersDisclaimers Why these schools?Why these schools?
These were NOT These were NOT necessarily the best necessarily the best undergraduate programs by undergraduate programs by any meansany means
We’re from big schools, We’re from big schools, and we wanted to find out and we wanted to find out what our “cousins” were what our “cousins” were doingdoing
Our results may be Our results may be buggy!buggy! We often have results We often have results from only one person, from only one person, with one perspective on with one perspective on the whole!the whole!
Other instructors may do Other instructors may do things differentlythings differently
6/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Gathering Data Wasn’t Gathering Data Wasn’t Easy!Easy!
MITMIT ““What you're looking What you're looking to do is indeed to do is indeed potentially very potentially very worthwhile.worthwhile.
However, after However, after discussing your discussing your request, including request, including with the department with the department heads, heads, we have we have unfortunately had to unfortunately had to conclude that we're conclude that we're not in a position to not in a position to respond at this respond at this timetime.”.”
YaleYale ““We have no We have no interest in interest in completing your completing your surveysurvey, which , which would take a great would take a great deal of effort for deal of effort for very little very little apparent gain.”apparent gain.”
People answered the survey with varying degree of detail and
completeness. Sometimes responses were opinions of one person, other times of several faculty.
7/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Thanks so much to…Thanks so much to… UC Berkeley:UC Berkeley: Mike Clancy & Brian Mike Clancy & Brian
HarveyHarvey CMU:CMU: Rich Pattis, Tom Cortina & Don Rich Pattis, Tom Cortina & Don
SlaterSlater Columbia: Columbia: Adam CannonAdam Cannon Georgia Tech: Georgia Tech: Mark GuzdialMark Guzdial Maryland:Maryland: Bill Gasarch & Jeff Foster Bill Gasarch & Jeff Foster NYU:NYU: Victor Shoup Victor Shoup Rice:Rice: John Greiner John Greiner Rutgers:Rutgers: Rich Martin, Sesh Venugopal Rich Martin, Sesh Venugopal
& Don Smith& Don Smith Stanford:Stanford: Nick Parlante & Julie Nick Parlante & Julie
ZelenskiZelenski UC Irvine:UC Irvine: David Kay David Kay UCLA:UCLA: David Smallberg David Smallberg UIUC:UIUC: Sam Kamin & Cinda Heeren Sam Kamin & Cinda Heeren USC:USC: Claire Bono Claire Bono UW:UW: Stuart Reges Stuart Reges Wisconsin:Wisconsin: Jim Skrentny & Perry Jim Skrentny & Perry
KivolowitzKivolowitz
Duke:Duke: Owen Astrachan & Camelia Owen Astrachan & Camelia EavesEaves
Brown:Brown: Chad Jenkins & Tom Chad Jenkins & Tom DoeppnerDoeppner
Caltech:Caltech: Leonard Schulman Leonard Schulman Cornell: Cornell: David GriesDavid Gries Harvard: Harvard: Matt Welsh & Mike SmithMatt Welsh & Mike Smith Johns Hopkins:Johns Hopkins: Rao Kosaraju, Rao Kosaraju,
Joanne HoulahanJoanne Houlahan Princeton:Princeton: Kevin Wayne Kevin Wayne Purdue:Purdue: Jeff Vitter & Buster Jeff Vitter & Buster
DunsmoreDunsmore UCSD:UCSD: Beth Simon & Paul Kube Beth Simon & Paul Kube UMass:UMass: Andrew McCallum Andrew McCallum UNC:UNC: Ketan Mayer-Patel Ketan Mayer-Patel UPenn:UPenn: C.J. Taylor C.J. Taylor UT Austin: UT Austin: Mike Dahlin & Roger Mike Dahlin & Roger
PriebePriebe
Jeremy HuddlestonJeremy Huddleston Samantha Jones, Samantha Jones, Brandon Johnson,Brandon Johnson,Beth TrushkowskyBeth Trushkowsky
8/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
InstitutionInstitution
Quarters vs. Quarters vs. semesters?semesters?
Offerings?Offerings? Flexibility?Flexibility? Types?Types?
9/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Is the institution on quarters or Is the institution on quarters or semesters?semesters?
Some schools on the Some schools on the west coast are on west coast are on quartersquarters
Opportunities for Opportunities for more varied classes more varied classes and tracks?and tracks?
Comparisons between Comparisons between courses can be trickycourses can be tricky
Type of term
6
26
QuarterSemester
10/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Course offeringsCourse offerings
Is a literacy course available Is a literacy course available for students not interested in for students not interested in programming?programming?
ModelsModels Fluency in Information TechnologyFluency in Information Technology Great Ideas in Computer ScienceGreat Ideas in Computer Science UPenn:UPenn: Networked Nature of Networked Nature of SocietySociety
UNC:UNC: Power Tools for the Mind Power Tools for the Mind
YesYes 2121
NoNo 77
11/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What is the flexibility in the lower-What is the flexibility in the lower-division sequence?division sequence?
Georgia Tech’s Threads Georgia Tech’s Threads curriculum provides a curriculum provides a great deal of great deal of flexibilityflexibility
Multiple introductions Multiple introductions to CSto CS
Brown (Java vs. Brown (Java vs. Scheme/ML-> Java)Scheme/ML-> Java)
Columbia (C/Java/Matlab)Columbia (C/Java/Matlab) UCI: multiple majors and UCI: multiple majors and
sequencessequences Inflexible curriculaInflexible curricula
Do not accept APDo not accept AP Required to take CS Required to take CS
courses and prereqs in courses and prereqs in first semesterfirst semester
10 schools noted that 10 schools noted that courses are part of a courses are part of a common first-yearcommon first-year
1
6
14
7EnormousModerateSomeNone
12/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What types?What types?
EthicsEthics 2 schools had 2 schools had
required Ethics required Ethics coursescourses
SurveysSurveys UCLA & Harvard have UCLA & Harvard have
survey as CS 1survey as CS 1 Irvine & UCSD have Irvine & UCSD have
courses where profs courses where profs talk about their worktalk about their work
How important is ethics?
Essential (2)
Minor (8)
None (11)
Important (2)
Survey courseavailable?
1510
4 Yes
No
Non-MajorsOnly
13/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
StaffStaff
Who are Who are instructors?instructors?
Do they change?Do they change? Who covers Who covers section?section?
14/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Are the instructors research Are the instructors research faculty, teaching faculty, or faculty, teaching faculty, or
graduate students?graduate students? TeachingTeaching
USC, CMU, Duke, Maryland, USC, CMU, Duke, Maryland, Rice, Stanford, UNC, Rice, Stanford, UNC, UCSD, UT AustinUCSD, UT Austin
BothBoth UW, NYU, Cal, Cornell, UW, NYU, Cal, Cornell,
Princeton, Columbia, Princeton, Columbia, Georgia TechGeorgia Tech
Depends on CourseDepends on Course Wisconsin, U Penn, UCLA, Wisconsin, U Penn, UCLA,
Rutgers, UIUC, UC IrvineRutgers, UIUC, UC Irvine ResearchResearch
Harvard, PurdueHarvard, Purdue QuotesQuotes
UCSD: “In theory Research UCSD: “In theory Research Fac, Teaching Fac or Fac, Teaching Fac or Lecturers. In reality, Lecturers. In reality, the latter 2.”the latter 2.”
9
7
6
2Teaching
Both
Dependson courseResearch
15/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Are the courses taught by the same person every Are the courses taught by the same person every semester?semester?
QuotesQuotes Purdue: Purdue: “No, but the “No, but the
pool is small”pool is small” UW:UW: “Lecturers teach “Lecturers teach
the course often, but the course often, but a large pool of people a large pool of people rotate in”rotate in”
UIUC:UIUC: “Instructor-led “Instructor-led courses yes, research-courses yes, research-led no”led no”
YesYes Caltech, Cornell, Caltech, Cornell,
Harvard, USC, UT Harvard, USC, UT AustinAustin
Depends on CourseDepends on Course U Penn, UC Irvine, U Penn, UC Irvine,
UCLA, UIUC UCLA, UIUC
5
416
Yes
Dependson course
No, butgenerallyregulars
16/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Who teaches discussion sections / Who teaches discussion sections / recitations? Labs?recitations? Labs?
AnalysisAnalysis People said “TAs” -- People said “TAs” --
it was not clear it was not clear whether grad or whether grad or undergradundergrad
No sections or labsNo sections or labs UNCUNC
QuotesQuotes UCSD:UCSD: Closed: Faculty Closed: Faculty
assisted by tutor, assisted by tutor, Open: TAsOpen: TAs
98
1 Tas
Some Tas,someFacultyNo sections,discussions,labs
17/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
DemographicsDemographics
What are the What are the enrollment numbers enrollment numbers and trends?and trends?
Who takes the Who takes the courses?courses?
What kinds of What kinds of grades are grades are awarded?awarded?
18/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Enrollment SummaryEnrollment Summary
Long-term, recentLong-term, recent Down,up: 5Down,up: 5
down over the long down over the long term up in recent term up in recent yearsyears
Down: 8Down: 8 Down, level: 7Down, level: 7
down over the long down over the long term steady in term steady in recent yearsrecent years
Up: 7Up: 7 cmu, harvard, cmu, harvard, princeton, ucla, princeton, ucla, upenn, ut, uwupenn, ut, uw
5
87
7Down, up
Down
Down,levelUp
19/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
One Take on EnrollmentOne Take on Enrollment
UC Berkeley Computer Science Annual Enrollment, 2002-present
0
100
200
300
400
500
600
700
800
900
2002 2003 2004 2005 2006
CS3
CS61A
CS61B
CS61C
CS70
20/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What is the typical grade histogram What is the typical grade histogram for the course?for the course?
Upper end Upper end Percentage of Percentage of As range from As range from 75% at Brown to 75% at Brown to 15% at Purdue15% at Purdue
Grade Distribution
0
5
10
15
20
25
30
35
40
test
%
ABCDF
21/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
ContentContent
LanguagesLanguages TextbooksTextbooks Stale vs FreshStale vs Fresh What works and What works and what doesn’twhat doesn’t
22/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What versions of what languages are What versions of what languages are taught?taught?
All JavaAll Java CMU, Duke, Princeton, CMU, Duke, Princeton,
Purdue, UNC, UT Austin, Purdue, UNC, UT Austin, UW, Rutgers & UCSDUW, Rutgers & UCSD
No Java (C++ & MIPS)No Java (C++ & MIPS) UCLAUCLA
Java Java C C MarylandMaryland
Java Java C++ C++ Stanford (just enough), Stanford (just enough),
Wisconsin & USCWisconsin & USC Java Java Matlab Matlab
CornellCornell Java & Scheme/MLJava & Scheme/ML
UC Irvine (Java UC Irvine (Java Scheme)Scheme)
Brown (Scheme/ML Brown (Scheme/ML Java) Java)
Java and 1 other...
3
1
1
2
C++C onlyMatlabScheme
9
14
1All Java
SomeJavaNo Java
23/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What versions of what languages are What versions of what languages are taught?taught?
3: Matlab, Python, Java3: Matlab, Python, Java Georgia TechGeorgia Tech
3: Java, C, Python3: Java, C, Python U PennU Penn
4: C, Ruby, little C++ & 4: C, Ruby, little C++ & JavaJava
HarvardHarvard 4: Scheme, Java, 4: Scheme, Java,
Assembler, C99Assembler, C99 UC Berkeley (MIPS)UC Berkeley (MIPS) Rice (Patel & Patel’s asm)Rice (Patel & Patel’s asm)
4: Java, C, Matlab, Python4: Java, C, Matlab, Python ColumbiaColumbia
5: Java, C++, C, 5: Java, C++, C, Assembler, Variety Assembler, Variety (student’s choice)(student’s choice)
UIUCUIUC 7: Scheme, C, C++, Java, 7: Scheme, C, C++, Java,
Python, ocaml, haskellPython, ocaml, haskell CaltechCaltech
Java and >2 others...
2
4
1
1
3457
24/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What are the textbooks What are the textbooks used?used?
BrownBrown Sanders and van DamSanders and van Dam
CalCal Grillmeyer, SICP, Grillmeyer, SICP,
Bates & Sierra, Bates & Sierra, Goodrich & TammassiaGoodrich & Tammassia, , K&R, K&R, Patterson & Patterson & HennesseyHennessey
CaltechCaltech SICP, Darnell & SICP, Darnell &
Margolis, K&RMargolis, K&R CMUCMU
Varies, sometimes Varies, sometimes nonenone
CornellCornell Gries & GriesGries & Gries
DukeDuke Goodrich & TammassiaGoodrich & Tammassia
Georgia TechGeorgia Tech Smith, GuzdialSmith, Guzdial
HarvardHarvard K&RK&R
Johns HopkinsJohns Hopkins Malik, Malik, Goodrich & Goodrich &
TammassiaTammassia, Deitel & , Deitel & DeitelDeitel
MarylandMaryland Lewis & Loftus, Bates and Lewis & Loftus, Bates and
Sierra, Koffman & Sierra, Koffman & Wolfgang, Sestoft, Reek, Wolfgang, Sestoft, Reek, Bryant & O’Hallaron, EppBryant & O’Hallaron, Epp
PrincetonPrinceton Sedgewick, Sedgewick & Sedgewick, Sedgewick &
WayneWayne PurduePurdue
SavitchSavitch
25/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What are the textbooks What are the textbooks used?used?
RiceRice Felleisen et all, Felleisen et all,
Rosen, Patt & Patel, Rosen, Patt & Patel, Bryant & O’HallaronBryant & O’Hallaron
RutgersRutgers Downey, Venugopal, Downey, Venugopal,
HorstmannHorstmann StanfordStanford
RobertsRoberts UC IrvineUC Irvine
Felliesen, Horstmann, Felliesen, Horstmann, Goodrich & TamassiaGoodrich & Tamassia
UCLAUCLA Savitch, CarraSavitch, Carrañoño
UCSDUCSD SavitchSavitch
UIUCUIUC Kamin et al, Rosen, Kamin et al, Rosen,
Woodbury, Collins, Woodbury, Collins, Mano, Mano, Patterson & Patterson & HennesseyHennessey, Stallings, , Stallings, McConnellMcConnell
UNCUNC Bloch, Metsker & WakeBloch, Metsker & Wake
USCUSC Forouzan & Gilberg, Forouzan & Gilberg,
Malik, Lauesen, LiangMalik, Lauesen, Liang UT AustinUT Austin
WeissWeiss UWUW
Reges & SteppReges & Stepp
26/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Is there a feeling that the course is Is there a feeling that the course is “fresh” or “stale”?“fresh” or “stale”?
LessonsLessons Some said no, but the Q Some said no, but the Q
was meant to evaluate a was meant to evaluate a range.range.
Q should have asked for a Q should have asked for a #!#!
Hard to admit staleness!Hard to admit staleness! QuotesQuotes
Wisc:Wisc: “I believe, and I “I believe, and I sense my students would sense my students would agree, that I teach the agree, that I teach the course in a manner that course in a manner that makes it feel fresh makes it feel fresh regardless of whether the regardless of whether the same content has been same content has been taught for ages.”taught for ages.”
UIUC:UIUC: “Not fresh, but “Not fresh, but little perceived need to little perceived need to change”change”
Fresh? Stale? Fresh
Fresh &NeutralNeutral
Fresh &StaleStale
Some Stale?Some Stale? Cal, Cornell, UIUCCal, Cornell, UIUC
““Probaby stale”Probaby stale” UCSDUCSD
““Going to discuss overhaul”Going to discuss overhaul” USCUSC
27/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What works really What works really wellwell and what and what is is brokenbroken??
BrownBrown High TA-student ratioHigh TA-student ratio Intro Experience Intro Experience
DisparityDisparity CalCal
Closed required labsClosed required labs Undergrad teaching staffUndergrad teaching staff More pair programmingMore pair programming
CalTechCalTech CS11 is several CS11 is several
minicourses.minicourses. Each student only takes 1 Each student only takes 1
of these a term (of their of these a term (of their choice)choice)
A course can be repeated A course can be repeated three times for creditthree times for credit
CMUCMU All courses taught in All courses taught in
small sections – no large small sections – no large lectures!lectures!
CornellCornell DrJava demos during DrJava demos during
lecturelecture Closed required labsClosed required labs 1/2 Matlab, 1/2 Java1/2 Matlab, 1/2 Java
Georgia TechGeorgia Tech (Compared to peers) very (Compared to peers) very
low attritionlow attrition 90% success in Media CS290% success in Media CS2
No closed labsNo closed labs HarvardHarvard
Teacher trainingTeacher training PrincetonPrinceton
Appl to commerce, Appl to commerce, science, engineering science, engineering w/real datasetsw/real datasets
Too much to cover!Too much to cover! Huge increase in Huge increase in
enrollmentsenrollments
28/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What works really What works really wellwell and what and what is is brokenbroken??
PurduePurdue Lecture, labLecture, lab Hard to have creative Hard to have creative
projects but consistent projects but consistent gradinggrading
RiceRice Some want to move intro Some want to move intro
Scheme course to JavaScheme course to Java Easier later transitionEasier later transition But, wide background But, wide background
variety!variety! RutgersRutgers
Online examsOnline exams Java as intro language!Java as intro language!
StanfordStanford Undergrad section leadersUndergrad section leaders Staffing in computer Staffing in computer
cluster while students work cluster while students work on assignments, which are on assignments, which are goodgood
We (all of us) still tuning We (all of us) still tuning CS1 JavaCS1 Java
UC IrvineUC Irvine Informatics curriculumInformatics curriculum
UCLAUCLA Getting students to Getting students to
program to specsprogram to specs Intro Experience disparityIntro Experience disparity
UCSDUCSD Accelerated intro course Accelerated intro course
tracktrack Closed labsClosed labs Teaching memory allocation Teaching memory allocation
to beginners to beginners UIUCUIUC
All courses!All courses! Undergrad-led reviewsUndergrad-led reviews Assembler EOS competitionAssembler EOS competition Programming studio Programming studio
structurestructure Low-level understandingLow-level understanding New student communitiesNew student communities
29/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What works really What works really wellwell and what and what is is brokenbroken??
U PennU Penn Compelling programming Compelling programming
assignmentsassignments USCUSC
Intro experience disparityIntro experience disparity UT AustinUT Austin
Students move well through Students move well through CS2CS2
Java doesn’t allow them to Java doesn’t allow them to learn memory management!learn memory management!
One course is “geeked out” One course is “geeked out” which could turn off non-which could turn off non-programmersprogrammers
UWUW Everything is fineEverything is fine
U WisconsinU Wisconsin Infornation overloadInfornation overload Reduce sources of infoReduce sources of info Active and collaborative Active and collaborative
learning!learning!
““Information overload has been Information overload has been mentioned repeatedly to me by mentioned repeatedly to me by students as a principle problem students as a principle problem in many of their courses. in many of their courses.
It is not uncommon for courses It is not uncommon for courses today to have a textbook or today to have a textbook or two, course notes, as well as two, course notes, as well as lecture slides which overwhelm lecture slides which overwhelm students with presentations of students with presentations of the same or similar material.the same or similar material.
Add to that course email, web Add to that course email, web pages/links, blogs, etc. and pages/links, blogs, etc. and students get trapped spending students get trapped spending too much time reading and too much time reading and reviewing and insufficient time reviewing and insufficient time actively applying their newly actively applying their newly gained knowledge.gained knowledge.
What has worked well for us is What has worked well for us is to reduce the sources of to reduce the sources of information and to direct information and to direct students’ efforts to more students’ efforts to more active and collaborative active and collaborative learning both in and out of learning both in and out of class.”class.”
30/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Delivery & StyleDelivery & Style
How are courses How are courses run?run?
What style(s) of What style(s) of programming are programming are taught?taught?
31/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Contact hoursContact hours
Standard seems to be 3 hours Standard seems to be 3 hours of lecture and 1-2 hours of of lecture and 1-2 hours of lab/discussion lab/discussion
What drives labs ?What drives labs ?
Some open labsSome open labsDiscussion sectionsDiscussion sections
ProblemProblem 2121
InstructInstructoror
77
32/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What programming paradigms are What programming paradigms are covered?covered?
Everyone includes some OOEveryone includes some OO ParadigmsParadigms
P: ProceduralP: Procedural O: Object-orientedO: Object-oriented F: FunctionalF: Functional L: LogicalL: Logical D: DeclarativeD: Declarative
OO is everywhereOO is everywhere About half say they teach About half say they teach
“objects-first”“objects-first”
Programming Paradigms
P,OP,O,F,LP,O,F,DP,O,FOVaries
33/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
MetaMeta
Grading Grading guideline?guideline?
What questions What questions were missing?were missing?
Anything else?Anything else? What are the What are the innovative innovative techniques being techniques being deployed?deployed?
34/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Is there an institutional or departmental Is there an institutional or departmental grading guideline?grading guideline?
Who answered yes?Who answered yes? Princeton:Princeton: 35% As 35% As UC Berkeley:UC Berkeley: 2.7 2.7 GPAGPA
UC San Diego:UC San Diego: 2.8 2.8 GPAGPA
HarvardHarvard Other answers:Other answers:
UIUC:UIUC: Grading Grading standard of standard of previous semester previous semester sent aroundsent around
20
4
NoYes
35/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What questions are missing on What questions are missing on this survey?this survey?
Brown: Brown: Are numbers at day 1 Are numbers at day 1 or N? or N?
Caltech:Caltech: Ask more about Ask more about languageslanguages
Us: scheme, java, then what Us: scheme, java, then what the student wants!the student wants!
CMU:CMU: What are your ugrads What are your ugrads paid?paid?
Cornell:Cornell: % of your incoming % of your incoming kids who have prog kids who have prog experience? (60%)experience? (60%)
Harvard:Harvard: Training of teaching Training of teaching staff?staff?
Princeton:Princeton: Demography of Demography of course should include majorcourse should include major
Rice:Rice: What related courses What related courses exist outside yoru dept? Are exist outside yoru dept? Are courses weeders? (us: no)courses weeders? (us: no)
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
36/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Was there anything not captured by this survey you Was there anything not captured by this survey you would like to add about your institution, staff, would like to add about your institution, staff,
demographics, content, delivery or style?demographics, content, delivery or style?
Brown: Brown: We don’t track We don’t track ethnicityethnicity
Caltech: Caltech: We have outstanding We have outstanding students. We try to give students. We try to give them resources they need to them resources they need to excel, & introduce material excel, & introduce material not found elsewherenot found elsewhere
Functional programming, Functional programming, ocaml, haskellocaml, haskell
Cornell:Cornell: We give CS or IS We give CS or IS majors in three colleges and majors in three colleges and have two different intro have two different intro courses (OO-first Java & courses (OO-first Java & Matlab)Matlab)
Harvard:Harvard: We work hard to We work hard to have gender-balanced (or all have gender-balanced (or all male) sections. This yr, > male) sections. This yr, > 1/2 of our teaching staff 1/2 of our teaching staff was female. We will soon was female. We will soon offer on our own survey of offer on our own survey of what works best to revamp what works best to revamp our intro courses!our intro courses!
37/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What are the innovative techniques being What are the innovative techniques being deployed?deployed?
MITMIT Overhauling EECS curriculumOverhauling EECS curriculum
Had been unchanged for Had been unchanged for nearly 30 yearsnearly 30 years
6.001 (Scheme) out, 6.001 (Scheme) out, Python Python in!in!
6.01 and 6.02 to be new 6.01 and 6.02 to be new mandatory intro coursesmandatory intro courses Offer intro to deptOffer intro to dept Cut broadly across EECSCut broadly across EECS Strong hands-on componentStrong hands-on component Build robots, apply Build robots, apply
concepts like radio, concepts like radio, Fourier transform, Fourier transform, networks, large systemsnetworks, large systems
Then choose 3 or 4 subjects Then choose 3 or 4 subjects from foundation coursesfrom foundation courses E.g., Circuits, Signals, E.g., Circuits, Signals,
SystemsSystems
www.eecs.mit.edu/images/EECS_VI_Fall06_FINAL.pdf www-tech.mit.edu/V126/N63/63course6.html
38/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What are the innovative techniques being What are the innovative techniques being deployed?deployed?
Georgia TechGeorgia Tech Computational Media Computational Media
degree!degree! Different entry points to Different entry points to
majormajor CS1 for different majorsCS1 for different majors
RoboticsRobotics Media ComputationMedia Computation
FlexibilityFlexibility CS Majors can take any CS1CS Majors can take any CS1
ContextualizationContextualization CS1s are typically dry!CS1s are typically dry! Students today use flikr, Students today use flikr,
youtube, etc -- leverage youtube, etc -- leverage that!that!
Allow the data to be THEIRSAllow the data to be THEIRS movies, sounds, pixmovies, sounds, pix
www.cm.gatech.edu
Wil Justis, Milton HSWil Justis, Milton HS
39/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
What are the innovative techniques being What are the innovative techniques being deployed?deployed?
UC BerkeleyUC Berkeley UC-WISEUC-WISE
Web-based learning systemWeb-based learning system Interactive, collaborative Interactive, collaborative
& assessment& assessment Decrease passive learning Decrease passive learning
in lecture (Old: 3 Lec, 2 in lecture (Old: 3 Lec, 2 Lab)Lab)
Increase active learning in Increase active learning in lab (New: 1 Lec, 6 Lab)lab (New: 1 Lec, 6 Lab)
Top-down instructionTop-down instruction Scheme, Java, C, AssemblerScheme, Java, C, Assembler
Self-paced coursesSelf-paced courses 1-unit course for programmers 1-unit course for programmers
in Matlab, Pascal, C, Scheme, in Matlab, Pascal, C, Scheme, Unix, C++, Java, PythonUnix, C++, Java, Python
Intro to programming in SchemeIntro to programming in Scheme Awesome undergrad tutors!Awesome undergrad tutors!
inst.eecs.berkeley.edu/~selfpace www.ucwise.org
40/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
Meta-lessons we Meta-lessons we learned?learned?
Too many questions!Too many questions! Should have limited Should have limited scope.scope.
Just first course?Just first course? Could have presented all Could have presented all results in one sessionresults in one session
Should be answerable in Should be answerable in minutes, not hoursminutes, not hours
Some questions redundantSome questions redundant Discrete math courses Discrete math courses fell through cracksfell through cracks
We leaned on We leaned on friendships to get friendships to get datadata Thank you!!Thank you!! Next time: get funding, Next time: get funding, provide incentivesprovide incentives
41/41A Survey of Introductory Computer Science Curricula www.cs.duke.edu/csed/openwiki/doku.php?
id=teaching:start
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
SIGCSE2007
ConclusionConclusion There is so much more There is so much more to squeeze out of the to squeeze out of the data!data! It’s all online!! We It’s all online!! We encourage you to squeeze!encourage you to squeeze!
There are clear There are clear clusters based on clusters based on language (& book) language (& book) choices, tracks for the choices, tracks for the major, etc.major, etc.
Best & Worst practicesBest & Worst practices There is some great There is some great innovation going on innovation going on around the country!around the country! MIT, GaTech, UC Berkeley, MIT, GaTech, UC Berkeley, Caltech, U Penn, …Caltech, U Penn, …