194 JPJO 7 (2) (2022) 194-201 Jurnal Pendidikan Jasmani dan Olahraga Available online at: https://ejournal.upi.edu/index.php/penjas/article/view/45021 DOI: https://doi.org/10.17509/jpjo.v7i2.45021 Cognitive Analysis Using Rasch Modeling As an Assessment for Physical Fitness Knowledge for College Students Didin Budiman, Ricky Wibowo Universitas Pendidikan Indonesia,Indonesia Article Info Article History : Received March 2022 Revised April 2022 Accepted Mei 2022 Available online September 2022 Keywords : Instrument Analysis, Physical Fitness Knowledge, Rasch Measurement Abstract Having good physical fitness is the foundation for carrying out daily activities. Main- taining physical fitness requires knowledge and understanding for performing mean- ingful physical activities. This study aimed to analyze the physical fitness knowledge test instrument for college students using the Rasch modelling data analysis technique. The design, development, and research method employed the product model and re- search tool that consisted of four stages, including analysis, design, development, and evaluation. The samples involved in the instrument trial were 70 students from one university in Indonesia. The analysis used was the content validity ratio for testing content validity and the item response theory/Rasch model test employing Winstep software to determine the validity and reliability of the test instrument. The results of the data analysis showed a match between the students and the test instruments with very good quality items. It concludes that the test instrument development using the Rasch model can detect student misconceptions on physical fitness knowledge materi- al.  Correspondence Address : Jl. Dr. Setiabudi No.229, Kota Bandung, Jawa Barat 40154 E-mail : dinbudiman1974@gmail.com https://ejournal.upi.edu/index.php/penjas/index 195 INTRODUCTION Health and physical fitness development has be- come a concern in the last few decades. Interest in de- veloping physical fitness results from expertise that strongly links physical activity and life expectancy (Gunter et al., 2012; Baird et al., 2019: Faienza et al., 2020). Although there are still estimates on the intensity and duration of physical activity for physical fitness, some experts agree that various forms of physical activ- ity will benefit the body (Sun et al., 2020; Soriano- Maldonado et al., 2022). For example, regular physical activity is the best way to avoid cardiovascular disease and lower blood pressure during submaximal exercise (Cheng et al., 2018; Lacombe et al., 2019; LaCroix et al., 2019). In addition, lifestyle can increase passion for life and reduce mortality. Physical fitness is also associ- ated with a reduced risk of coronary heart disease, dia- betes, obesity and osteoporosis (Wei et al., 2000; Katzmarzyk et al., 2007; Liao et al., 2013; Babiuch et al., 2021). On the other hand, good physical fitness can increase cardiovascular endurance, muscle strength, endurance and flexibility (Wilmore & Knuttgen, 2003; Sillanpää et al., 2008; Gäbler et al., 2018; ). Given the clear benefits of health-related physical fitness concerns, more need to increase physical fitness knowledge levels. This raises much knowledge about health-related physical fitness from practitioners, expert questions and educators. With attention to exercise, the body's fitness determines whether they do an exercise based on what they know. One of the goals of physical education at the uni- versity level is to provide knowledge and skills and maintain fitness levels so that students can apply life- style habits (Budiman et al., 2018). For students to de- velop these habits, the realm of physical fitness related to health needs to be directed (Corbin, 1987; Mirzaev, A. M., 2021). Body composition, flexibility, cardiovas- cular endurance, strength endurance and strength are part of physical fitness related to health. An understand- ing of physical fitness related to health and skills can be achieved by incorporating goals and objectives in train- ing that benefit from physical activity to achieve a healthy and active lifestyle. In some European coun- tries, inculcating an active lifestyle is an important mis- sion of physical education (Almond & Harris, 1998; Cardon & De Bourdeaudhuij, 2002; Harris, 2005), and there is an excellent emphasis on health-related activi- ties related activities in the physical education curricu- lum (Almond & Harris, 1998; Harris, 2005). Promoting a healthy and active lifestyle is a significant goal in physical education (UNESCO, 2015). Regular physical activity is an important lifestyle component for both children and adults. There have been many programs to change healthy lifestyles based on social cognitive the- ory (Bandura, 2004; Budiman et al., 2018; Rovniak et al., 2022) and behaviour change based on self- regulation models (Kanfer & Gaelick-Buys, 1991; López -Gil et al., 2020). In addition, researchers have identified some specific skills, such as self- management, that can play a role in helping to change lifestyles (Cardon et al., 2009). From previous research, there is already Health- related fitness knowledge (HRFK) which includes the concepts and skills needed to improve and maintain levels of fitness and physical activity (Corbin CB, Lind- sey, 1988; Society of Health and Physical Educators, 2014). HRFK is a crucial element for health and physi- cal education to guide one's ability to control one's movements so that they are physically independent and plan their fitness activities (Corbin CB, Lindsey, 1988; Society of Health and Physical Educators, 2014). HRFK is also believed to be a critical component in increasing students' awareness of healthy living behav- iour (Demetriou et al., 2015). Other studies also state that HRFK strongly correlates with exercise intensity and physical activity (Ferguson KJ et al., 1989; Haslem L, 2016). Along with the increasing awareness of the im- portance of physical fitness for a person, there is also a growing need for tests that can determine the amount of knowledge about the practice desired and mastered by students; This research was conducted based on this view. With an understanding of physical fitness, stu- dents' ability to choose the most desired exercise in their physical fitness will increase. Physical fitness knowledge tests are expected to reflect students' habits, knowledge, and experiences in physical activity. Since physical education is an integral part of habituation to a healthy and active lifestyle (Harris, J., 2014), students should know not only games and activities but also the general psychological and physiological implications of various types of physical activity. To determine the ef- fectiveness of learning physical education, sports and health, educators must have several measuring tools to Copyright © 2022, authors, e-ISSN : 2580-071X , p-ISSN : 2085-6180 Didin Budiman & Ricky Wibowo / Jurnal Pendidikan Jasmani dan Olahraga 7 (2) (2022) 196 determine the level of student knowledge and whether they can apply their knowledge in practice. This test will assist educators and students in determining the desired practical knowledge in physical fitness. Based on the description above, understanding physical fitness is fundamental to staying active in their daily lives, especially for adults. Previous research found instruments to evaluate the understanding of physical fitness in middle and high schools. However, it is limited to understanding Health-related fitness knowledge. While understanding the Skills-related fit- ness knowledge, the authors have not found the results of studies investigating these fitness components. The purpose of the research developed in this study was to develop Physical Fitness Knowledge items for students. The questions compiled are the content of teaching ma- terials related to physical fitness to develop and pro- mote a healthy and active lifestyle. METHODS The research method used is Design, Develop- ment, Research (Richey & Klein, 2007) with the cate- gory Product model and research tool; this model is product development. The stages in the Product and tool research model are; Analysis, design, development, and evaluation. The variable to be studied is the stu- dent's ability, as seen from the computer-based test re- sults for the Final Semester Examination in the Physical Education class. The instrument of the study is in the form of items using a dichotomous model in the form of multiple choices. Data was collected through a comput- er-based test in the Physical Education class. Before the computer-based test in the physical ed- ucation class, content validity and face validity were tested. Face validity is checked by checking the content of the questions in the form of the validity of the word or sentence so that it has a clear understanding and is not misinterpreted on each test item. The instrument has good face validity if the instrument is easy to under- stand and students do not experience difficulties an- swering questions. Content validity is carried out in two stages: (1) determining the content of the definition used and (b) developing indicators that cover all the things contained in the definition. Content validity and face validity in this study were carried out by asking for experts who were competent in their fields and were the material given to students. Furthermore, the computer-based test questions in the physical education class were tested on students who took this PE class. The test results data were ana- lyzed to determine the questions' characteristics empiri- cally. The approach used in the data analysis of the test results is the item response theory/Rasch model with the help of the Winstep software. Validity and Reliability Analysis To test the validity of the items, the Rasch was used based on the following criteria Model (Sumintono & Widhiarso, 2015): Accepted Outfit Mean Square value (MNSQ): 0,5 MNSQ <1,5 Accepted Z-Standard Outfit Value (ZTSD) : -2,0 < 2STD < + 2,0 Measure Correlation of Valued Point (Pt Mean Corr) : 0,4 < Pt Mean Corr < 0,85 If a computer-based test item in the physical edu- cation class meets at least the two criteria above, then the item is declared usable; in other words, the item is valid. Reliability is the consistency of test results. The measurement results must be the same (relatively the same) if the measurements are given to the same sub- ject, even though different people carry them out at dif- ferent times and places. . Data Source The types of data collected in this study are classi- fied into two, namely : a.Primary data Data on students taking physical education classes that meet the sample inclusion requirements is obtained through computer-based tets physical education classes using the items that have been provided. b.Secondary data Data on the list of the number of students who are taking physical education class Data Analysis Technique The test insrument data was analyzed using the Rasch item response theory approach using the WIN- STEPS program. The steps taken are : Copyright © 2022, authors, e-ISSN : 2580-071X , p-ISSN : 2085-6180 Didin Budiman & Ricky Wibowo / Jurnal Pendidikan Jasmani dan Olahraga 7 (2) (2022) 197 1. Assumption test a. Unidimensional Assumption Test The unidimensional assumption test aims to deter- mine whether the test is proven only to measure one dominant dimension. The unidimensional assumption test is carried out by performing factor analysis. b. Local Independence Assumption Test The local independence assumption test is carried out to determine if the ability to answer the test questions conducted by students is not influenced by whether or not the answers from other test participants are correct. The local independence assumption test is met if the covariance value is close to zero. c. Parameter Invariance Assumption Test The parameter invariance assumption test was conducted to determine the consistency of the charac- teristics of the items answered even though the items varied. The test results are carried out by looking at the scree plot. The parameter invariance assumption is achieved if the plot spreads and approaches the normal line. 2. Item Fit to Model The item fits the model if it has an outfit Mean Square (MNSQ) value of 0.5 < MNSQ <1.5 and the Point Measure Correlation (PT Mean Corr) value is not negative. RESULT This research was carried out at the University of Education Indonesia, and data collection was in August 2021. The research subjects are students taking physical education classes according to the inclusion and exclu- sion criteria set. Before taking the data, a study of the material related to physical fitness knowledge was car- ried out. The material analyzed is the curriculum mate- rial for physical education and sports. Furthermore, the questions are by the level of thinking ability based on Bloom's theory. The preparation of the difficulty level is divided into 25 questions consisting of 5% or 1 ques- tion C1, 5% or 2 questions C2, 45% or 12 questions C3, 25% or 6 questions C4, 10% or 3 questions C5 and 5% or 1 question C6. The distribution of questions can be seen in table 1. Twenty-five items were then tested on students. Sixty-seven students are the samples in this study. The following is data from an objective test of 25 questions for students taking physical education and sports clas- ses. To examine the assessment result sheet to find sug- gestions for improvement related to aspects of material content with the suitability of problem-solving indica- tors and problem-solving indicators with question indi- cators. Furthermore, data analysis was carried out to determine the level of validity and reliability of the in- strument. Furthermore, the computer-based test questions in the Physical Education class were tested on students taking physical education classes. The test results data were analyzed to determine the questions' characteris- tics empirically. The approach used in the data analysis of the test results is the Response Theory/Rasch Model with the Winstep software. The reliability analysis results can be seen using the Winsteps program in the Summary Statistics table. The validity results can be seen and analyzed with the Winsteps program in the Outfit table to see the suitabil- ity of the items that function in the normal category to be used as a measurement of student misconceptions by requirements in table 2. Characteristics of Items and Respondents The analysis using the Winsteps program provides information, both in terms of items and respondents, showing differences in the items and students analyzed using the Rasch model, indicating the occurrence of misconceptions for some students. According to Arikunto (2019), a question that can be answered cor- rectly by smart students and students who are less intel- ligent is a question that is not good because it does not have distinguishing power. The description of the distri- bution of the ability of 67 students and the distribution of the difficulty of the items on the same scale. The re- sults of data analysis in Figure 1 show that the average student is in medium ability with a logit value of +1 (011, 021, 031, 040, 041, 011, 021, 031, 040, 041, 010, 013, 029, 033, 036, 048, and there are also two students with low abilities with a logit value of -1 which means that there are still some students who still have miscon- ceptions about the concept of the material being tested. From Figure 1, we can also see that two questions have a high difficulty value, namely, questions num- Copyright © 2022, authors, e-ISSN : 2580-071X , p-ISSN : 2085-6180 Didin Budiman & Ricky Wibowo / Jurnal Pendidikan Jasmani dan Olahraga 7 (2) (2022) 198 bered Q19, Q24, and Q16, and also, most of the stu- dents understand the concepts listed in questions Q19, Q24, and Q16. Some questions have a medium difficul- ty value, such as questions Q6, Q12, Q11 & Q2, with a logit value between -0 and +0, and some have a low level of difficulty with a logit value below -1. Logit below -1 is what must be revised again. This result is by Boopathiraj and Chellamani (2013), who say that ques- tions that have high discriminating power are questions that students who have low test scores cannot answer the questions correctly. The previous explanation proves that some students still have misconceptions about physical fitness material. From the analysis results, it was found that the average value of all students working on the questions given was 2.5 logit. This value is greater than the aver- age difficulty value, meaning that there is a tendency for students to be higher than the difficulty level of the question. For example, students get a value of 0.58, and the question is obtained at 4.08. Therefore, it can be said that the greater the value of separation, the greater the quality of the instrument used is very good. This is because it can identify groups of questions with groups of respondents. Item Validity The data analysis results show that there is a valid level on the instrument construct of the questions devel- oped to obtain a level of conformity between student responses and the test instrument. The following table Copyright © 2022, authors, e-ISSN : 2580-071X , p-ISSN : 2085-6180 No Material C1 C2 C3 C4 C5 C6 TOTAL 1 Active and healthy lifestyle 1 1 2 Benefits health and active lifestyle 1 1 3 Components of health and skills related fitness 1 1 2 4 Physical activity and classification of physical activity levels 2 2 5 Physical fitness test (adult fitness test; cardiorespiratory, muscular strength, muscular strength endurance, flexibility, body composi- tion) 1 1 6 Endurance activities 1 1 7 Rest, exercise and maximum heart-rate 1 1 8 Body Mass Index 1 1 2 9 muscular endurance activities 1 1 10 Benefits of warming-up and cooling down 1 1 2 11 Types and functions of carbohydrates, fats, proteins, vitamins, minerals, water 1 1 12 muscular strength activities 1 1 2 13 Fleksibilitas: static, dynamic and PNF 1 1 14 Calories and physical activity 1 1 2 15 Components of Skills- related fitness : Agility, speed and reaction 2 2 16 Principles of fitness activity program 1 1 · Overload · Progression · Spesificity · Reversibility · Recovery · Individual 17 F.I.T.T formula in fitness activities 1 1 2 - Frequency - Intensity - Time - Type TOTAL 1 2 12 6 3 1 25 Table 1. The composition of the difficulty level Question Table 2. Item Validity Criteria No Reference Limit Value 1 OutfitMean Square (MNSQ) 0,5 < MNSQ < 1,5 2 Outfi Z-Standard (ZSTD) -2,0 < ZSTD < +2,0 3 Point Measure Correlation (Pt Mean Corr) 0,4 < Pt Mean Corr < 0,85 Didin Budiman & Ricky Wibowo / Jurnal Pendidikan Jasmani dan Olahraga 7 (2) (2022) 199 shows how the items developed are normal or not in measuring students' misconceptions about physical fit- ness material. The requirement is to know whether the items can be categorized as acceptable or not by looking at the MNSQ scale with a range of 0.5 < MNSQ < 1.5. If you look at Figure 2, it shows that of the 25 items devel- oped, they are in the good or accepted category, so it can be concluded that there are no misconceptions from students about these items. The ZSTD value scale is categorized as acceptable or not with a range of -2.0