In need of a Base Case
Comments about coding, research and life

Ok, first a short reflective blog post, just in case I’m still a part of people’s RSS feeds!

This past year the blogging has significantly declined because I have shifted my focus from reading about others work to conducting some of my own. Here are some of the projects I’m working on right now:

* Naive models of Looping and Instructional Benefit: I am currently getting my feet wet in exploring conceptual change in a fairly constrained domain looking at for and while structures and mapping people’s naive predilection for one type of loop syntax over another yields benefits in learning gains. I’ve constructed a small cognitive tutor and using a within subjects crossed design to explore the question. Data still incoming..

* Misconceptions in Software Engineering: I’ve completed a project with a SE PhD student looking at the misconceptions students at CMU have in regards to software engineering and offering a novel assessment methodology that measures not only the misconceptions but their strength as well.

* Animation vs. Static Instruction in young children with the day/night cycle: This is an interdisciplinary project I am completing with a psychology student where we are looking at the effect of static (still images) vs. animated instruction in 2nd and 3rd graders with regards to conceptual change and misconception dispersal.

I’m now starting to think about the direction for my thesis research and to that effect am starting to read again. Expect a number of posts about some talks from last week, as well as conversations I have had and readings I am doing.

Well, heres to a good start to what I hope will be a productive summer!

In preparation for writing the introduction/motivation section of a research proposal I’m reflecting on some previous work in assessing naive understanding of computing concepts. One of the articles recommended to me by Sally Fincher at ICER was the Commonsense Computing series. The first paper in that series looked at students naive concepts of sorting.

First let me say I approve of the work and most of the methodology that was applied. I’m in the middle of my own process for writing research proposals which are a combination of drawing on similar work and also exposing weaknesses in that work with motivate your own work. This blog is not meant to be critical, just observant of some of the things in the paper I feel were important for me to think about.

First of all, their selection of students (first day of a computer programming class) may already introduce a bias into the student’s answers. Because students have already enrolled and showed up to a class which is generally understood to exist in a computational setting, that may have shifted the student’s perceptions in that direction. Is that necessarily a bad thing? I dont know, after all we teach them these concepts in a computational setting. One of the things I’m hoping to explore in my own work to help provide a measurement of how much the questions get answered by “what they think is the right answer, or what the teacher wants to hear” is the idea of attachment strength. If anyone knows of an article that talks about attachment strength to a model I am looking for a good citation for the paper. (as well as a little help in designing that part of the assessment)

The second to last paragraph in the introduction states “The results suggest that beginners can describe algorithms, but their models of the machine and instructions differ from those of many instructors. In particular, the results suggest that instructors should guide students to understand a virtual machine in which numbers are primitive objects …” This lead to a note in the margin that reads: Are the students’ models “wrong” or just abstracted at a different level? Why should we guide them towards a NEW model rather than providing insight into their own?

The paper states in a couple of places that one of the goals of this work is to inform better instruction based upon the findings, but the largest instructional change recommended was to shift from while loops to do-until loop structures. Any intuitions expressed by the student that appear to represent a different level of abstraction were labeled as a misconception and it was recommended that the instructor work to move the student to the instructors way of thinking. While this is probably the correct response in most cases, I’m wondering if it is right in all. Are there cases when a misconception is based upon a different level of abstraction where we should simply introduce students to the concept at their abstraction level and then progress to a deeper one over time? I guess this is one of the broader questions I hope to address in my work. hm.. Any comments would be much appreciated. Even if they are based on naive models

Earlier this morning Marcia Linn gave the Keynote at ICER entitled “Learning to Teach Computer Programming”. The work that she talked about, while containing some historical perspective about teaching computer science, was mostly about a new report “Fostering Learning in the Networked World: The Cyberlearning Opportunity and Challenge” and two initiatives: Computational Thinking and 21st Century Skills Movement.

I have not read the Cyberlearning report, so I do not have a lot to comment about it.

As far as the Computational Thinking and 21st Century Skills movement - first I was very happy to hear the “21st Century Skills” agenda introduced at a computer science. She even gave a link to the 21stcenturyskills.org website and showed their “rainbow” curriculum model.

Marcia showed us a simulation from the WISE collection of Science simulations and tried to model how this was a computational thinking/21st Century Skill activity. (It was about global temperature and you could control the amount of C02 that was added to the environment) I was not convinced that it was truly a computational thinking activity. One of the features of computational thinking that I was struck by the first time I heard Jeanette Wing speak about it was the idea that Automation was one of the three key aspects of computational thinking. Its not just about looking a representation of information, but it is about somehow automating some process. The WISE collection of activities is great, but I’m not sure its really computational thinking.

Marcia also talked about a cycle of knowledge building that can be used through a tutor or electronic environment where students go through a 4 stage process of generating ideas, adding ideas, evaluating those ideas and finally sorting the ideas based on the evaluation. This reminded me a little of the misconception research that says you need to expose student’s misconceptions in order to move past them, however it was unclear how incorrect ideas in this process would be “weeded out”.

Still processing what my take away from that talk will be.

The CSTA and others have been talking a lot about certification as their new report “Ensuring Exemplary Teaching in an Essential Discipline”. The question is as we talk about certification in CS I would caution us against relying too heavily on models of certification for other disciplines.

There was an interesting article a short time ago in Curriculum Matters (an Education Week blog) about some of the identified problems with the teaching profession, especially in STEM disciplines. Also at this year’s IES (Institute for Education Sciences) conference I attended a panel on teacher certification (they were comparing alternative certification against traditional certification, and showed no significant difference in student performance between the two.

As we move computer science forward and hope to better prepare our CS teachers for the classroom, lets make sure that the preparations themselves are right.

Recently there has been an ongoing thread on the SIGCSE list serv about whether or not there is a need for a CS Education PhD. The conversation has gone in multiple directions, some arguing that everyone needs an education degree, and talk about K12 teachers. Here’s my reply:

I have been lurking on this thread for a couple of days, and I would like to weigh in.

I am currently pursuing a self-defined PhD in Computer Science Education from Carnegie Mellon University. The reason it is self defined is because there is no formal program for this. I am building on a core of cognitive science work and computer science fundamentals (since CMU is granting me the degree through the CS department they want to make sure I am qualified).

Before starting this degree I taught K12 for 10 years, and I do not believe that this degree is about making me a better teacher. I do not believe that you need a degree in CS Education to be a good CS teacher at the college or HS level. I do however believe that we need people with CS Education PhD’s to help us shape the curriculum and understand how our students learn.

I have a lot of respect for the work that the SIGCSE community does, however the more work I do in this degree the more I come to understand that education research, especially at the cognitive science level is HARD to do well. In our community we have a large number of CS researchers who are exceptional educators and are doing good research, however there needs to be more work done about the underlying cognitive processes that support the learning of CS.

Not everyone needs a CS PhD’s to be an exceptional teacher, however we do need some CS Ed PhD’s in order to help understand our students, our curriculum and WHY some of our teachers are so exceptional so that we can begin to replicate the small bubbles of excellence that we see throughout high schools and CS departments all over.

The ACM recently commissioned a marketing research report to find out what kind of message would be most successful in changing the attitudes of today’s youth towards computer science as a potential career path, and hopefully at the same time encouraging more people to enroll in computer science degree programs.

The report, which can be found here details the research methodology as well as the results.

In addition to the message findings, one of the more intriguing findings to me was that while there was significant difference for gender, there was no significant difference between ethnicities. African American and Latino boys were just as likely as Caucasian boys to think that computing is a viable and interesting career choice. This supports the research that Jane Margolis did in Stuck in the Shallow End, which indicated that inequities in the educational system were a large factor in why certain populations were missing from the computer science landscape.

In terms of messaging, the message that was most successful with the women was that Computing empowers you to do good. 38% of women reported this message as “Very Compelling”, and it was the third best ranked message overall.

Something in the key results that echoes a comment I made during a session at SIGCSE was that “The strongest positive driver towards computer science, or an openness to a career in computing is ‘having the power to create and discover new things’.” I made the comment that so many of our assignments are about recreating programs that students can find faster on Google than they can even load up the IDE and begin to think about programming it. Again I recommend to the community that we use this information to continue to design assignments that make use of new technology and social patterns within the use of computing in order to inspire our students and engage them in our classes.

For my learning and motivation class the paper assigned for tomorrow is “The Four-Phase Model of Interest Development”. For anyone concerned with the declining enrollments in CS or who wants to engage underrepresented groups in CS I would recommend reading it. The paper is less of a description of original research, and more of a proposal of a framework for developing interest.

So here are some of the highlights and my comments from the paper:

First of all interest is defined as “a motivational variable [that] refers to the psychological state of engaging or the predisposition to reengage with particular classes of objects, events, or ideas over time.” The thing that struck me was the idea that by definition interest = action. Later the author states that “it is a biological function of the psychological state of interest in the sense that the person is engaged physically, cognitively, or symbolically with the object of his or her interest”. So the question becomes not how do I get my students to find this interesting, but instead how do I inspire action in them. Its a different question and one that leads to different strategies I think.

One example that they gave in the paper was of a girl named Julia who picked up a magazine in a doctor’s office and based on a description in a magazine. The authors talk about interest being biologically grounded in a seeking behavior. This leads me to ask do our students need to be somehow primed into this seeking behavior? either by activating an already existing need or by offering them a scenario that prompts them to internalize a new need.

Finally, a key aspect to having students develop the most long term and pervasive type of interest is the fostering of a questioning mentality in the student. Its discussed in several parts of the paper - here are some examples:

An essential component of the four phase model is that support and opportunities to pursue interest-related questions are necessary for each phase of interest. Without these, regression to a previous phase of interest can be expected to occur.

In later phases of interest development, as a person begins to generate curiosity questions, he or she seeks repeated engagement and has not only positive feelings but also increased stored knowledge and stored value for particular content.

a person’s developing understanding of particular activities or ideas and the generation of curiosity questions. The process of pursuing answers to curiosity questions, for example, is accompanied by positive feelings that surface in anticipation of and work with particular content as well as feelings generated in present engagement.

However, as individual interest begins to emerge in the late phases, it is important that students also be encouraged to generate their own questions. Students need models of people seriously engaging with the questions of a discipline. For students’ interests to continue to develop, however, they also need to generate their own curiosity questions to connect their present understandings to alternate perspectives.

When was the last time a lesson, an activity or an assignment prompted the question “I wonder how they do that” either in you or your students?

Rethinking this in broader terms, looking at a lot of efforts to engage students in computer science many of the topics have been about showing students how something works. How a robot works, how a program works, etc. Can we make an intervention that instead prompts students to be curious? to ask how? and to see themselves as the kind of person who could look for those answers? Maybe thats why CSI became such a popular show and eventually prompted a societal response - not because the show was “cool” but because it constantly offered situations where professionals modeled interacting with the domain, asking difficult questions and engaging in problem solving. It prompted curiosity that fueled long term individual interest.

What do you think?

So I am reading Frank Pajares’ paper entitled Self-Efficacy Beliefs in Academic Settings for a very interesting course I am taking on Learning and Motivation.

Pajares states:

Researchers have reported that the mathematics self-efficacy of college undergraduates is more predictive of their mathematics interest and choice of math-related courses and majors than either their prior math achievement or math outcome expectations and that male undergraduates report higher mathematics self-efficacy than do female undergraduates

This is really interesting to me. Other places in the paper it makes reference to the fact that in general men have higher self-efficacy than women. These two things combined and the perception of computer science, engineering and physics as very mathematically intensive courses of study may be contributors to our low perceptions. I think it would be really interesting to run a large scale assessment of undergraduate self efficacy (across a number of universities) and see if student’s self efficacy for simple computing tasks also correlates to math.

Theres lots of other interesting things in this paper as well as one by Barry Zimmerman entitled Self-Efficacy: An Essential Motive to Learn I highly recommend them to anyone doing research (or considering it) into why there is such an ethnic and gender disparity in computer science enrollment.

So, in the process of reading every paper on mental models I can find in preparation for a study this spring, I have become a little concerned with the nature of just about every work I can get my hands on about mental models in computer science.
It seems to me that every paper is about predicting success in a computer science course, or how certain questions can predict success or failure on an exam.
While these are worthwhile research questions to be asking, shouldnt we be asking the other half as well? What existing models do our students have, what models do they form over the course of the semester and how do we influence those models in such a way as to not predict success, but foster it?
This is what my research is about - how can we measure what they already know (hopefully correctly) and then either build upon that or expose the faulty models in order to produce successful students.
I am currently working on how to assess what models they do hold. More info as that thought process works itself out.

A mental model is conventionally understood as an internal representation of an external event. Mental models have been used in the past to understand complex interactions between humans and the world around them.

They are important to my research because I want to know how people think about and interact with their computer programs as a way to inform introductory computer science education.

Many of the papers I have been reading recently (Ramalingam especially) have used some measure to assess the “mental models” of students, but to me it seems more like a measurement of their ability to chunk and remember.

My current dilemma is how do I (without doing an in depth interview) come up with a general way to look at mental models of programming. Any thoughts?