Eye movements dataset for objective-based assessment of object-oriented programming knowledge

The Eye Movements dataset for Objective Assessment contains eye gaze data and learners' scores in an objective-based object-oriented programming assessment. The learners’ knowledge was assessed for various programming concepts like object initialisation, variable declaration, constructors etc. The eye gaze data of learners were collected using the commercial eye tracker, and their responses using the Think Aloud method. The eye gaze data was then mapped to the contextual information in the stimulus, which included questions, keywords, and answer options. The raw and analysed data are available to learn learners' cognitive behavior during an objective assessment.


Subject
Computer Science Education Specific subject area Eye-tracking Data for Learner Assessment in Object-Oriented Programming Type of data Table How the data were acquired Data was gathered by eye-tracking using an SMI Redn 60 Hz Professional eye tracker device.

Data format Raw Data Processed Data Description of data collection
The experimental study is conducted in the eye-tracking research lab in the Department of Computer Science and Engineering, Amrita School of Computing, with 15 learners.All learners have prior programming knowledge and are undergoing an introductory course in Java Programming.The experiment was designed in the SMI Experimental Suite 360 ֯ .The remote eye-tracking device was mounted on the laptop at a distance of 60-70 cm.The responses of the learners were collected using the Think Aloud method.Data source location • Institution: Amrita Vishwa Vidyapeetham

Value of the Data
The Eye Movements dataset for Objective Assessment (EMOA) consists of learners' demographic information and eye gaze data collected for an empirical study to understand their cognitive behavior during an objective-based assessment in an object-oriented programming language.
• The dataset can be used to infer the learner's problem-solving process and cognitive traits.
• The task designed in the empirical study can be used as a template to construct new tasks to analyze learners' problem-solving processes at different levels of thinking as described by Bloom's Taxonomy.• The design can also be extended to conduct similar empirical studies to understand learner knowledge in different courses.

Objective
The EMOA dataset was collected as no dataset is available in the literature to understand learners' problem-solving processes and cognitive traits during the objective assessment for programming language.The objective assessment was based on the lower level of Bloom's Taxonomy [1] called Remembrance on topics like keywords, operators, decision-making, loops, constructors, object creation, and initialization of object-oriented programming paradigm.The dataset was analyzed to get insights into cognitive traits like confidence level, quickness in task completion, task engagement, inattentional blindness to critical information, and wavering behavior [2] .The dataset is available for researchers to analyze various aspects of learners in an objective-based assessment.It can also be used in creating machine learning models to predict learner performance and behavior on similar tasks.

Data Description
The EMOA dataset consists of three files-EMOADemographics.csv,EMOAEventData.csv, and EMOASemanticData.csv.The EMOADemographics.csv consists of the learner's information about age, gender, prior knowledge in programming, and level of education pursued.EMOAEvent-Data.csvcontains learners' event data like total fixation, saccade, and blink data generated using SMI Redn Professional BeGaze Software.Table 1 represents the participants' demographic information, and Table 2 represents the event data description.The file EMOASemanticData.csvconsists of the semantic gaze information of learners on various Area of Interest(AOI).Each Multiple Choice Question (MCQs) comprises six AOI regions-Question, Option A, Option B, Option C, Option D and Keyword, as shown in Fig. 1 .The Question is the MCQ question; Option A, Option B, Option C and Option D are the various choices in the MCQ.Keyword is the key information in the question that helps the learner identify the right option in the MCQ.For each of these AOIs fixation duration and fixation count were computed using the open gaze analysis platform PyGaze [3] .These gaze information are called semantic features as they provide contextual information of what is viewed by the learner.Table 3 consists of the semantic features.

Experimental Design, Materials and Methods
The experimental design consists of the setup, stimulus design and participant recruitment.

Experimental setup
The experimental setup comprises hardware and software for data collection.The hardware system consists of a laptop, and a remote-mounted eye tracker, SMI Redn Professional, with

Table 3
Semantic Features as in EMOASemanticData.csv.60 Hz [4] .The eye tracker was mounted on the lower part of the laptop screen 60-70 cm away from the learner.The experiment was set up using SMI Experimental Suite 360 ° [5] .The dataset was processed and exported from SMI BeGaze Software [6] .The EMTADataset was further analyzed and the semantic features were then generated using in-house algorithms running on PyGaze platform.

Stimulus design
Five MCQs based on various object-oriented concepts were given as an assessment task.Each MCQ has a question, a keyword and four options.The keyword is the critical information in the question.All the MCQs were designed to assess the learners at a lower level of Bloom's Taxonomy called Remembrance.The following are the details of MCQs, the related concept, the key-words that provide valuable information towards the correct choice and various answer options as in the experimental study [2] : • MCQ1: Which keyword is used by method to refer to the object that invoked it?
The Keyword: object that invoked it Options
Identification.Identified as STUD-P01…….STUD-P15 Trial Trial Information.Identified as Trial001, Trial002,… Trial 005 Correctly Answered Response of the learner was correct(Yes) or incorrect (No) Total Fixation Number of Fixations in the Trial Total Duration Total Fixation Duration for the Trial in ms Q

Table 1
Demographic information of participants as in EMOADemographics.csv.

Table 2
Event Data Description of EMEventData.csv.