Friday, June 12, 2015

Our work is accepted at SIGCSE 2017!

We're pleased to share that our work's been accepted at SIGCSE 2017. ACM SIGCSE is a premier conference focusing on computer science education. Started in 1970, it's a fun conference where each year, tons of new ideas get published and discussed by educators and computer scientists. We believed that the principles we followed to define our fun tutorials should be shared with this community. There were a lot of nuances we considered to ensure that the tutorials ended up being fun, hands-on and engaging while ensuring that kids weren't cognitively burdened. This was a new way to teach kids this topic, which otherwise is taught in undergraduate courses.

This year's submission process was fairly elaborate. Five-seven reviewers went through each submission. A total of 300 papers were submitted this year, of which nearly 100 were selected. The reviewers were extremely pleased with our work, agreeing that this was definitely a template which educators could try at their schools.

A PDF of the paper can be downloaded from this page - http://research.aspiringminds.com/publications

The reviewers' detailed comments have been mentioned below -

----------------------- REVIEW 1 ---------------------
PAPER: 68
TITLE: Introducing Data Science to School Kids
AUTHORS: Shashank Srikant and Varun Aggarwal

OVERALL EVALUATION: 5 (Clear Accept: Content, presentation, and writing meet professional norms; improvements may be advisable but acceptable as is)

----------- Summary -----------
Data Science is an important field of study as industry increasingly need to employ Data Scientist. Marketing and creating awareness of Data Science as a career opportunity to students and scholars is important. This paper proposes and evaluates a tutorial that is presented to scholars/students completing Grades 5-9 that allows then to gain hands-on exposure to the field of Data Science.

----------- Strengths -----------
The idea is novice. Other institutions can use this tutorial as a basis to develop similar interventions to expose young adults to the exciting field of Data Science and possible career opportunities. The tutorial is well motivated and discussed, allowing young scholars/students to identify possible friends by exposing them to the complete process a data scientist will generally follow and finally make decisions.

----------- OVERALL EVALUATION -----------
The idea and tutorial is new and provides the basis for similar studies at other institutions, promoting careers in Computer Science. Data Science is a new career opportunity and Higher Education Institutions are increasingly implementing Data Science programmes. The tutorial is a new novice approach and presents a well thought-through case study. A detailed discussion of the tutorial is provided, including motivation for certain decisions and providing hands-on experience to scholars/students. The authors' provide a detailed motivation for their choice of technologies and data, including the evaluation and discussion of the results.

Personally, I do not recommend the author's write a paper in the first person (We, I, Us). I'm further not sure but are Grade 5's Secondary School? Secondary school grades are generally Grade 8-12? Further, in the Abstract the authors' indicate that they "limited the pre-requisites for the kids to the knowledge of counting, addition, percentages and comparisons". Generalising that this knowledge allows scholars/students in different countries to use spread sheets easily is not true, for example in Africa limited scholars in Grade 5 has ever been exposed to computer technologies.

I suggest the authors review the paper critically to ensure a scientific writing style. Table 4 -the general scientific presentation of a 5 point Likert Scale is fro Strongly Disagree (1) to Strongly Agree (5). Finally the authors must indicate if the scholars/students learned or in the future will consider a career as a Data Scientist. This was the objective of the exercise.


----------------------- REVIEW 2 ---------------------

OVERALL EVALUATION: 4 (Marginal Tend to Accept: Content has merit, but accuracy, clarity, completeness, and/or writing should and could be improved in time)

----------- Summary -----------
This paper describes how school children in grades 5 - 9 were given a half day tutorial in data science in several cities. Efforts were made to minimise prerequisite knowledge,to  maximise engagement and avoid the need  to use complicated tools. The authors list their design principles for the hands-on exercise and the workflow of the tutorial as well as some feedback from the children participating. Less than 5% of the children did not find it interesting, according to the authors. Being able to build predictions and see if they worked on real data seemed to hold the children's attention.

Those colleagues who teach Data Science at undergraduate level may be interested in the lessons learned from this novel exercise of making the subject interesting to children.

----------- Strengths -----------
Novel nature of taking a subject , normally taught at undergraduate level, and engaging much younger students.
The authors' design principles used.

----------- OVERALL EVALUATION -----------
Data Science has not yet become standard material on many Computing Science courses. This paper may encourage conference attendees to consider its inclusion, based on what has been achieved with much younger students in less than a day. Others may be encouraged to use this approach to do 'outreach' and encourage potential students to consider enrolling for Computing related degrees and could challenge some misconceptions about Computing.


----------------------- REVIEW 3 ---------------------

OVERALL EVALUATION: 4 (Marginal Tend to Accept: Content has merit, but accuracy, clarity, completeness, and/or writing should and could be improved in time)

----------- Summary -----------
The submission reports on a half-day data science organised for school children, and presents design principles for creating such an exercise. It will be of interest to teachers who want to introduce data science ideas to young students

----------- Strengths -----------
The paper clearly describes the principles on which the tutorial was based, and gives a detailed account of the way the tutorial was conducted. This would be very helpful for anyone who wants to implement a similar activity.

----------- OVERALL EVALUATION -----------
The paper gives a detailed account of the design principles and conduct of the tutorial, from which it would be easy for other teachers to create similar activities. In fact, it appears that the supporting materials for this particular exercise will be made widely available. The design is strongly justified and this appears to be an interesting exercise that was well received. There is a mention of students being asked to blog about what they had learned, in addition to a questionnaire - was this separate from the "subjective comments" that are mentioned also? In some places the paper is a bit verbose or repetitive, though, for example much of table 1 repeats points made in the text.


Recommendations:
State the age range of the students - "5th to 9th grades" may bot be meaningful for all of the audience
Replace the word "kids" with a more formal term
"We also interchangeably refer to the dependent variables as output variables and independent variables as input variables respectively" - choose one set of terminology and stick to it
Table 2 - put data under the correct headings
Figure 2 - three parts in one figure, separate into different figures or clearly label a,b and c on figure


----------------------- REVIEW 4 ---------------------

OVERALL EVALUATION: 4 (Marginal Tend to Accept: Content has merit, but accuracy, clarity, completeness, and/or writing should and could be improved in time)

----------- Summary -----------
This is an interesting paper which presents the authors half day tutorial for pupils in grades 5 to 9 introducing Data Science. The approach appears to engage the pupils through, what seems, a fun and practical hands on experience. The authors have attempted to keep the activity highly visual and fun. This paper should appeal to teachers and the authors appear to be offering their work as a template for development. The authors have attempted to cover various aspects of Data science including gathering, process and visualising.

----------- Strengths -----------
I think the strength of this paper is the overall description of how the half day tutorial was conceived and implemented. The undoubted enthusiasm that the authors show throughout the paper for the subject area and the honesty with which they write.

----------- OVERALL EVALUATION -----------
The authors are clearly passionate about the subject area and have strived to create a tutorial that will cover the major aspects of data science in a fun and interactive way for the participants. They have attempted to keep the level of knowledge required by the participants as low as possible while still delivering a meaningful learning experience. On page 2 the symbol in front of each of the numbered sections should be replaced with the word section and the frequent use of the word "kid" be replaced with a more suitable term such as "pupil". I think the paper would be of interest to teachers and worth including at the conference.


----------------------- REVIEW 5 ---------------------

OVERALL EVALUATION: 5 (Clear Accept: Content, presentation, and writing meet professional norms; improvements may be advisable but acceptable as is)

----------- Summary -----------
Authors organized a half-day long data science tutorial for kids in grades 5 through 9. Their aim was to expose them to the full cycle of a typical supervised learning approach - data collection, data entry, data visualization, feature engineering, model building, model testing and data permissions. In the paper, they discuss the design choices made while developing the dataset, the method
and the pedagogy for the tutorial.

----------- Strengths -----------
The approach draws from different pedagogic theories like experiential learning, problem-based learning, cooperative learning, cognitive apprenticeship, and blended learning; hence the design is theoretically grounded.

----------- OVERALL EVALUATION -----------
Text under the title "2 Design Consideration" does not match  the information in Table 1, hence this is confusing. Other than that the writing is good, and the material is validated with data from 4 different contexts.


-------------------------  METAREVIEW  ------------------------
PAPER: 68
TITLE: Introducing Data Science to School Kids
AUTHORS: Shashank Srikant and Varun Aggarwal

The authors provide a nice case study of the design of an introduction to data science, aimed at students in grades 5-9. The treatment is novel; the activities are interesting and fun. The overly informal presentation (which used words such as "kids") drew some criticism from reviewers. Other small flaws include wordiness and bugs in table setup.

No comments:

Post a Comment