The Phoenix MetaLanguage Teaching Programming by Classroom Debate'

1
The Phoenix Meta-LanguageTeaching Programming
by Classroom Debate.

• Raymond Flood Bob Lockhart
• Department for Continuing Education
• The University of Oxford.
• 1 Wellington Square
• Oxford, OX1 2JA
• Wednesday, September 1st.
• Fifth annual LTSN-ICS conference
• University of Ulster, Jordanstown

2
The Programming Debate

• Programming central to modern computing courses.
• Poses difficulties as an academic topic.
• Most people feel it should be taught
  non-passively.
• Debate centres on which high-level language and
  what support facilities.
• In this talk, we will discuss our approach.

3
We shall

• Describe the nature of our courses and students.
• Outline our approach.
• Explain the facilities offered by the Phoenix
  tool.
• Illustrate our use of it in teaching.
• Present student responses to it.
• Share some of our resources.

4
One course - two delivery modes

• Undergraduate diploma in computing.
• Part time study over two years.
• 120 CATS points at level two.
• Twelve topics - basic undergraduate computer
  science.
• Programming approx 20.
• Continuous assessment plus an annual examination.
• No formal admission requirements.

5
Class course

• Two-hour lecture weekly - October to May.
• Three weekend schools each year.
• One exam each year, in June.
• Single class of 15-20 students.
• Variety of lecturers.
• Majority of students have no previous programming
  experience.
• Possession of a computer is not essential.
• Limited amount of hands-on computing.

6
Our approach

• Concepts of low-level programming.
• Phoenix programming-language generator.
• The 4 D approach!
• Debate facilities.
• Decide what they should be.
• Discover the consequences!
• Destination, Java!

7
Bare Bones

• Brookshears universal language exemplar.
• ONE of the dialects offered by Phoenix.
• Stripped-down programming language.
• Single control structure.
• Only variable type is non-negative integer.
• I/O undefined.
• Many important questions unresolved.

8
Illustration of Bare Bones

• clear Z
• while X not 0 do
• incr Z
• decr X
• end

9
The questions

• Case sensitive?
• Pre-set variables?
• Decrementing zero values?
• Annotation?
• Students can decide on all these issues.
• They reach a consensus.
• We implement their decisions.
• They discover the consequences.

10
How Phoenix works

• A language generator written in Java.
• Particular dialects correspond to settings of
  internal state variables.
• Students take the decisions but instructors
  implement them.
• Once variables are set, running Phoenix produces
  this code window which students use to program.

11
The programming interface
12
The student experience

• Phoenix is only used at the start of the course.
• First year we have attempted this! Only 11
  students.
• Reaction obtained by questionnaire and
  observation. (See http//www.conted.ox.ac.uk/cleat
  / for details).
• Only one student with previous programming
  experience.

13
Questionnaire

• Only one student did not enjoy programming.
• Only one reported the programming easy!
• Mixed signals about preferring Java only!
• All but two students were positive on Bare Bones.
• Mixed signals might relate to when the
  questionnaire was administered and the perceived
  utility of Java.

14
Specifics

• 1. I feel capable of writing a small Java
• program now.
• 2. I feel I understand the terms type,
• declaration, variable, assignment, control
• structure, loop, procedure, parameter.
• 3. I feel I understand the terms class, object,
  message constructor, attribute, behaviour,
  function.

15
Specifics

• 1. I feel capable of writing a small Java
• program now. 5 agreed, 2 neutral, 3
  disagreed.
• 2. I feel I understand the terms type,
• declaration, variable, assignment, control
• structure, loop, procedure, parameter.
• 3. I feel I understand the terms class, object,
  message constructor, attribute, behaviour,
  function.

16
Specifics

• 1. I feel capable of writing a small Java
• program now. 5 agreed, 2 neutral, 3
  disagreed.
• 2. I feel I understand the terms type,
• declaration, variable, assignment, control
• structure, loop, procedure, parameter.
• 2 agreed strongly, 4 agreed, 2 neutral, 2
  disagreed.
• 3. I feel I understand the terms class, object,
  message constructor, attribute, behaviour,
  function.

17
Specifics

• 1. I feel capable of writing a small Java
• program now. 5 agreed, 2 neutral, 3
  disagreed.
• 2. I feel I understand the terms type,
• declaration, variable, assignment, control
• structure, loop, procedure, parameter.
• 2 agreed strongly, 4 agreed, 2 neutral, 2
  disagreed.
• 3. I feel I understand the terms class, object,
  message constructor, attribute, behaviour,
  function.
• 6 agreed. 2 neutral, 2 disagreed.

18
Further points

• Students seemed more confident about concepts
  than had been the case in other years.
• Students welcomed further programming courses ( 2
  strongly agreed, 4 agreed, 4 neutral)
• The programming questions were among the most
  popular in their final exam.

19
Future expansion

• Local variables?
• Re-use and procedural units?
• More flexible control structures?
• Direct interface to instructor settings?
• Package as an Applet?
• Comparative studies?
• Phoenix version 2?

20
Acknowledgements

• John Axford Course Director, Internet diploma
  in computing.
• Suzie McMahon, for help with the programming
  questionnaire.
• Pete Thomas conversations on programming!

21
Thank You!
Raymond Flood
Bob Lockhart