Understanding students’ abstractions in block-based programming environments: A performance based evaluation

Highlights

• The study examined the effect of using block-based coding environments on enhancing secondary school students’ abstraction skills.

• A rubric was created to reveal the students’ abstraction performances.

• The students performed high in elimination, focusing and generalization; however, students’ performances were relatively low in customization.

• The nature of the problems, affordances of block-based programming environments and coding constructs were the factors affected the abstraction process.

Abstract

Providing computational problems for enhancing students’ abstraction skills and monitoring how students make abstractions is difficult in block-based programming environments (BBPEs). Thus, concrete examples and principles are needed to guide computer science teachers about understanding and enhancing students’ abstractions. This study aims to examine the effect of using block-based coding environments on enhancing secondary school students’ abstraction skills. Referring to the programming knowledge, a rubric was created to analyze the data from screen recordings, observation and interviews were used together to reveal the students’ abstraction performances. The results suggested that students performed high in elimination, focusing and generalization; however, students’ performances were relatively low in customization. Students’ explanations were mostly related the nature of the problems, affordances of BBPE and the programming constructs used in coding. We hope the study will provide insights for the efforts on instructional designs for successful abstraction experiences for young students.

Introduction

Teaching computational thinking (CT) to young audience is a highly debated issue and gained considerable interest among policymakers for introducing computer science in secondary school settings. Wing (2011) described CT as a high level skill that everyone should have such as reading, writing and basic mathematics. Various conceptualizations of CT were taken into consideration including algorithmic thinking, logical inquiry as well as strategies that involve problem solving strategies such as abstraction, decomposition, generalization, and collaboration (Apostolellis, Stewart, Frisina, & Kafura, 2014; Basawapatna, Repenning, Koh, & Savignano, 2014; Kalelioğlu, Gülbahar, & Kukul, 2016; Lee et al., 2011; Wing, 2008). Table 1 presents various definitions of CT with regard to the skills included in it.

Table 1 indicates that the most common skills in the CT conceptualizations are abstraction, algorithm thinking, automation, data collection, data analysis, data presentation, parsing, simultaneous work, pattern recognition, pattern generalization and modeling.

Algorithmic thinking is determining the processes to find the solution by showing them step by step in a problem. Automation can be considered as a labor-saving process where a digital system is used to perform a series of repetitive tasks quickly and efficiently (Lee et al., 2011). Pattern recognition explains the sequence of repetitive operations, observing patterns, and repetitive patterns in the data. Generalization is a way to solve new problems quickly based on previous solution experiences (Csizmadia et al., 2015). Decomposition is extracting a component consisting of complex or multiple structures. One leading skill in CT is abstraction which is considered as a fundamental skill for computer scientists. Abstraction is defined as ignoring less important situations by focusing on the features required to find the desired properties (CSTA, 2016). When defining CT, referring to thinking like a computer scientist (Wing, 2006) includes ability of doing abstraction. It is considered to purify from all situations except the property sought in an event or object (Gülbahar, Kert, & Kalelioğlu, 2019).

Gibson (2012) argued that high school is too late for exposing students to abstraction for the first time. In this sense, researchers argued that teaching abstraction in the computational context to young students, whose abilities of abstraction are not fully developed (Piaget, 1983), is a challenging tasks. However, teaching and learning abstraction is considered as challenging. This difficulty comes from the nature of abstraction that it is viewed as a fundamental idea (Armoni & Gal-Ezer, 2014).

Some efforts were provided to suggest various ways to teach abstraction to enhance students’ abstraction skills (Bucci, Long, & Weide, 2001; Çınar & Tüzün, 2020; Kramer, 2007). For example, Çınar and Tüzün (2020) in their study addressed that educational robot programming is as effective as traditional programming activities to result in positive outcomes in terms of abstraction. Similarly, Yesharim and Ben-Ari (2017) found that abstraction in simple algorithmic ideas and program design can be taught by using robots. On the other hand, use of BBPEs as a platform for teaching abstraction as a sub-skill of CT has become very prevalent.Given the importance of abstraction, teachers incorporate BBPEs into their lessons to enhance students’ abstraction skills. In this sense, evaluating the power of these environments in terms of abstraction may provide valuable insights for educators. Thus, investigating how BBPEs affect students’ abstraction skills may suggest some principles for instructional designers to design and use these environments better in the computer science courses.