Development of A Curriculum Type Preference Scale for Preschool Education Vocational University

Exploring students ’ preferences for curriculum types is beneficial for further designing curriculum structures based on the differentiated needs of students, thereby improving curriculum and education quality. To make up for the lack of an effective measurement instrument for investigating the curriculum type preferences of Chinese vocational university students majoring in preschool education, a 5-point Likert scale is designed and developed in this study. The development procedure consists of three phases: (1) the scale generation phase; (2) the scale evaluation phase, including expert evaluation and target population pretesting; and (3) the scale testing phase, including first testing, second testing, and third testing. The results show that the initial scale consisted of three constructs (theory curriculum, application curriculum, and internship curriculum) and 21 items in total. By removing one item from the initial scale, 20 items in the final scale have been proven to have acceptable validity and reliability. The conclusions suggest that the curriculum type preference scale can be applied in practice, and the scale development procedure and methods in this study can serve as a theoretical reference for scale developers.


Introduction
In China, it has become a trend for vocational universities to offer preschool education majors to cultivate practical early childhood education teachers (Hong, 2019).Correspondingly, a large number of students with different personality traits, learning levels, and educational needs have chosen to study preschool education in vocational universities, exhibiting significant individual differences (Hong, 2019).Curriculum is the main carrier for cultivating students in universities (Yuan, 2010;Gu & Shi, 2010), and dividing curriculum according to different standards forms curriculum types (Feng, 2007).For example, according to Zhu and Yi (2017), curriculum can be divided into social-based curriculum and child-based curriculum based on different educational philosophies; it can be divided into compulsory and optional curriculum based on different study requirements; and it can be divided into explicit Vol 13, Issue 1, (2024) E- ISSN: 2226-6348 To Link this Article: http://dx.doi.org/10.6007/IJARPED/v13-i1/20749DOI:10.6007/IJARPED/v13-i1/20749 Published Online: 16 February 2024 curriculum and implicit curriculum based on different curriculum expectations.In this context, understanding the curriculum type preferences of vocational university students majoring in preschool education is beneficial for further organizing curriculum types into curriculum structures in appropriate proportions according to the differentiated needs of students, thereby improving curriculum quality and better serving the cultivation of early childhood education talents (Zheng, 2018).
However, the existing curriculum structure of preschool education majors in Chinese vocational universities was too unified for all students and had not been designed based on their curriculum type preferences (Tao & Luo, 2020;Xu, 2020).In addition, theoretical research mainly investigated students' evaluation of the difficulty level of preschool education curriculum content, the rationality of curriculum quantity, the rationality of curriculum structure, and overall satisfaction with curriculum design in vocational universities, without conducting specialized research and investigation on students' preferences for curriculum types (Tao & Luo, 2020;Xu, 2020;Zheng, 2018).More importantly, in the research on curriculum surveys, interviews and qualitative analysis were mainly employed, while quantitative research only used non-scale questionnaires and descriptive statistical analysis, lacking a scale that has been rigorously validated as a measurement instrument (Tao & Luo, 2020;Wen et al., 2018;Xu, 2020).
Therefore, in order to fill the gap between ideal and reality, the purpose of this study is to develop a 5-point Likert scale of curriculum type preference for preschool education in vocational universities based on a systematic scale development procedure, as a measurement instrument to understand students' curriculum type preferences.

Methods
The scale development procedure in this study employed the three phases: scale generation, scale evaluation, and scale testing (Boateng et al., 2018;DeVellis & Thorpe, 2021;Kyriazos & Stalikas, 2018;Morgado, 2017).The scale generation phase is to develop an initial scale of the curriculum type preferences.The scale evaluation phase includes two steps: the first step is to utilize expert evaluation to judge the content validity of the scale; the second step is to utilize the target population pretesting to test the face validity of the scale.The scale testing phase includes three steps: the first step is to explore the construct validity of the scale through a first test; the second step is to verify the construct validity and measure the internal consistency of the scale through a second test; the third step is to check the test-retest reliability of the scale through a third test (Figure 1).

Phase One: Scale Generation
The objective of this phase is to develop an initial scale, including defining concepts, identifying constructs, and developing corresponding items (Carpenter, 2018;Kyriazos & Stalikas, 2018).The deductive method (literature review) and the inductive method (focus groups) were employed to generate an initial scale in this study (Boateng et al., 2018;Hinkin, 2005;Morgado, 2017).First, six relevant experts were invited to participate in a focus group as recommended by Creswell and Creswell (2018) (Table 1), and a semi-structured interview protocol was developed (Appendix 1).Then, the thematic analysis method was used to analyze interview data, and the analysis procedure was as follows: (1) familiarizing data; (2) generating codes; (3) searching, reviewing, and defining themes; and (4) forming the report (Braun & Clarke, 2006).

Phase Two: Scale Evaluation
Step 1: Expert Evaluation The objective of this step is to use the content validity index (CVI) (Yusoff, 2019) to measure the content validity of the scale.The procedure consists of six steps: (1) developing a content validation form (Appendix 2); (2) recruiting six experts for evaluating the relevance of items (Table 1); (3) implementing the content validity of the scale; (4) checking the domains and items of the scale; (5) providing scores for the items on the scale; and (6) calculating the CVI of the scale (Yusoff, 2019).
Step 2: Target Population Pretesting The objective of this step is to adopt cognitive interviewing (Howard, 2018) to pretest the participants to judge the face validity (e. g., wording concerns) of the scale.20 Chinese vocational university students majoring in preschool education were invited as participants as suggested by Lenzner et al (2016), and a semi-structured protocol of cognitive interviewing Collins (2003) was adapted in this study (Appendix 3).Moreover, a procedure of cognitive interviewing Drennan (2003) was used in this study, which includes two stages: in the concurrent interview stage, the think-aloud method is used to encourage students who are thinking about the items to verbalize their thoughts.In the retrospective interview stage, the probing technique is used to recall the memory of students who have completed the items to find existing problems.Finally, the thematic analysis method was used to analyze the data, and the analysis procedure was described in the scale generation phase.

Phase Three: Scale Testing
Step 1: First Testing The objective of this step is to use exploratory factor analysis (EFA) (Fabrigar & Wegener, 2011) to explore the construct of the scale by conducting the first test.Specifically, principal components analysis (PCA) was used to extract factors, and varimax was used as the rotational method Taherdoost et al (2022), and statistically analyzed through the Statistical Package for the Social Sciences (SPSS).In addition, 200 Chinese preschool education students at vocational universities were planned as respondents according to previous recommendations (Comrey & Lee, 1992;Johanson & Brooks, 2010).However, 210 questionnaires were distributed, and 207 valid questionnaires (182 females and 25 males) were recovered.
Step 2: Second Testing The objective of this step is to confirm the construct of the scale and check the internal consistency of the scale by conducting a second test.A fresh sample with 205 Chinese vocational university students majoring in preschool education (181 females and 24 males) was tested, and the data analysis process was as follows: firstly, confirmatory factor analysis (CFA) was used to verify the construct validity of the scale (Brown, 2015).Specifically, maximum likelihood (ML) was used as an estimator (Brown, 2015), and statistically analyzed using Mplus.Subsequently, Cronbach's alpha was applied to identify the internal consistency of the scale (Tavakol & Dennick, 2011), and SPSS was used to analyze the data.
Step 3: Third Testing The objective of this step is to measure the test-retest reliability (Guttman, 1945) of the scale by conducting a third test.This study conducted a third test in the third week after the second test and selected 100 students (88 females and 12 males) as respondents from the sample of the second test based on existing suggestions (Kennedy, 2022;Kurpius & Stafford, 2006).In addition, the intraclass correlation coefficient (ICC) was used to analyze the test-retest reliability of the scale (Koo & Li, 2016;Weir, 2005).Specifically, ICC estimates and their 95% confident intervals were calculated using SPSS based on a single-measurement, absoluteagreement, 2-way mixed-effects model (Koo & Li, 2016).

Initial Scale
On the one hand, through the literature review (Bu, 2005;Yang & Cao, 2009;Yang & Kang, 2006), it was found that based on the criterion of "the proportion of theory and practice", preschool education curriculum can be divided into three types: theory curriculum, application curriculum, and internship curriculum.The theory curriculum mainly focuses on the teaching of systematic theoretical knowledge in the form of lectures, discussions, questions, analysis, and reasoning.The application curriculum is a step-by-step simulation exercise and practice based on the theoretical guidance, which contains both theoretical proportion and practical proportion.The internship curriculum is a direct experience in kindergarten.The corresponding positions of the three types can be found in the dimension of " Theoretical-Practical" (Figure 2).On the other hand, three themes were refined through a focus group, and six experts unanimously believed that, in reality, preschool education curriculum in vocational universities can be divided into three types.
Theme 1: Theory Curriculum Six experts unanimously agreed that courses that mainly combine logical analysis and audiovisual methods as the main teaching and learning methods, with collective teaching as the main organizational form and theoretical knowledge as the teaching content are classified as theory curriculum."The course objectives require students to master systematic theoretical knowledge and be able to discuss and analyze theoretical problems such as Preschool Psychology, Preschool Pedagogy, and Preschool Hygiene, which we define as theory curriculum" (Expert 1).The literature review and expert discussion yielded the same finding.Therefore, this study divided the scale into three constructs: theory curriculum (TC), application curriculum (AC), and internship curriculum (IC).Afterwards, the corresponding seven items under each construct were developed, an item pool containing 21 items was generated, and a 5-point Likert initial scale was developed (Appendix 2).

Content Validity
When researchers measure the CVI, both the item-level content validity index (I-CVI) and scale-level content validity index (S-CVI) should be considered (Yusoff, 2019).According to the benchmark that the CVI estimates should be 0.83 or above with the participation of six evaluators (Polit & Beck, 2006;Polit et al., 2007), the result of the I-CVI showed that except for TC7 (I-CVI =0.5) and AC4 (I-CVI =0.67) (Table 2), the remaining items (I-CVI=1) had excellent content validity (Table 2).Other than that, the S-CVI/Ave (=0.96) and S-CVI/UA (=0.90) were higher than 0.83 (Table 2), which indicated that the content validity of the entire scale was acceptable.Afterwards, TC7 and AC4 were improved and certified based on expert suggestions.

Face Validity
The result of the pretesting showed that the target population had a clear understanding of the items and made choices, without any ambiguity, inappropriate wording, or redundant statements.Therefore, the original items were retained without any modifications.

Construct Validity -EFA
The EFA report needs to present the following three parts: (1) data inspection; (2) factor retention; and (3) factor interpretation (Howard, 2016).Firstly, the result of the data inspection showed that the value of Kaiser-Meyer-Olkin (KMO) was 0.907 (Table 3).According to the guideline for interpreting KMO values (.90 to 1.0 = marvelous; .80 to .89= meritorious; .70 to .79 = middling; .60 to .69 = mediocre; .50 to .59 = miserable; below .50= unacceptable) (Kaiser, 1974), this result indicated that sampling adequacy was excellent.Moreover, the result of Bartlett's test of Sphericity showed strong significance (p=0.000<0.01) (Bartlett, 1950;Tabachnick & Fidell, 2007) (Table 3), indicating that the data was suitable for factor analysis.Subsequently, the result of the factor retention showed that there were three factors with eigenvalues greater than one (Table 4), indicating that 21 items could be divided into three constructs (Kaiser, 1960).In addition, the cumulative percentage of variance (CPV) was 64.306% (Table 4), which met the standard of explained variance (commonly as low as 50-60% in the humanities) (Pett et al., 2003).Finally, according to Hair et al. (2009), the values of factor loading should be at least 0.5, and above 0.7 was better.Through comprehensive consideration, this study determined that the cut-off value of the EFA factor loading was at least 0.6.The result of factor interpretation showed that except for AC1, which had a factor loading value of 0.572, all other items had values greater than 0.6 (Table 5), indicating that the correlation between the remaining 20 items and the corresponding constructs met the standard, while AC1 needs to be removed.Note: TC1-TC7 = Theory Curriculum Item 1-7; AC1-AC7 = Application Curriculum Item 1-7; IC1-IC7 = Internship Curriculum Item 1-7.

Construct Validity -CFA
In terms of verifying the construct validity of the scale, model fit indices and standardized factor loadings are commonly used to explain the results of CFA (Jackson et al., 2009;Schreiber et al., 2006).Table 6 presents the output of Mplus for the model fit indices.
Figure 3 presents the output of Mplus for the standardized factor loadings.Consistent with the factor loading standard of EFA, this study determined that the cut-off value of the CFA factor loading was at least 0.6.The result showed that the factor loading values of all items were greater than 0.6, indicating that the correlation between all factors and the corresponding items was within the acceptable range.

Internal Consistency
The result of the internal consistency showed that the values of Cronbach's alpha of the whole scale and three subscales were between 0.851 and 0.929 (Table 7).Referring to the interpretation for Cronbach's alpha (≥.9 = Excellent; ≥.8 = Good; ≥.7 = Acceptable; ≥.6 = Questionable; ≥.5 = Poor; ≤.5 = Unacceptable) (George & Mallery, 2003), the result suggested that both the whole scale and subscales had at least good reliability.

Test-Retest Reliability
The result showed that the ICC estimate of all items ranged from 0.833 to 0.977, so the estimate of ICC with the 95% confidence interval of all items ranged from 0.762 to 0.985 (Table 8).Relying on ICC Interpretation (ICC < 0.5 = Poor; 0.5 ≤ ICC < 0.75 = Moderate; 0.75 ≤ ICC < 0.9 = Good; 0.9 ≤ ICC = Excellent) (Koo & Li, 2016;Portney & Watkins, 2000), the results showed that all items at least achieved good reliability.

Discussion
The findings from this study showed that by improving TC7, AC4, and removing AC1, a scale containing 20 items was developed and can be used in a specific context (Appendix 4).During the development process, validity and reliability were analyzed, and the following discussion focuses on these two sections: (Kelley, 1927).This study tested the content validity, face validity, and construct validity of the scale.Firstly, the results of the content validity showed that the correlation between TC7 (I-CVI=0.5)and AC4 (I-CVI=0.67)with their corresponding domains still needs improvement (Table 2).For the original statement of TC7 (Preschool Family Education), experts believed that some universities do not conduct this course purely as a theory curriculum, but rather incorporate a certain proportion of practice, such as designing situational performances.Therefore, experts suggested replacing this item with "Philosophy of Preschool Children's Education", which is more relevant to the domain of theory curriculum.For the original statement of AC4 (Preschool Child Care), experts believed that some universities integrated this course into "Preschool Hygiene" when carrying out this course.Therefore, experts suggested replacing this item with "Storytelling of Preschool Children", which is more relevant to the domain of application curriculum.
Moreover, the pretest on the target population found that the face validity of the scale was good, and respondents were able to clearly understand and make choices about the statement of the item.The reason is likely because the statement of all items is the familiar course name for the students, without ambiguity, inappropriate wording, or redundant statements.
Finally, from the EFA result, it can be clearly seen that the factor loading estimate of AC1 (0.572) was relatively low in the domain of the corresponding application curriculum, and also has a weak correlation in the theory curriculum (0.372) (Table 5).The possible reason for this is that the original item statement of AC1 is "Implementation of Kindergarten Education Activities", and this course content mainly designs kindergarten education activities and conducts corresponding simulation teaching based on theoretical knowledge, which belongs to the application curriculum.However, in reality, it has been observed that some lecturers who teach this course lack practical experience in kindergarten and the practical ability to guide students, leading to a large extent to teaching this course as a purely theoretical course, resulting in differences in students' positioning of this course.In addition, it was due to the strict screening of items using EFA (factor loading of 0.6 or above) and the removal of poorquality items that CFA verified good construct validity in the remaining 20 items.Bruton et al (2000) stated that reliability refers to the repeatability or consistency of a measuring test or quantitative research.This study considered two commonly measured reliabilities: internal consistency and test-retest reliability.The result showed that the whole scale and subscales had good to excellent internal consistency, and the test-retest reliability of all items also achieved good to excellent stability.The reason for this is that on the one hand, it reflects that the scale indeed has good reliability, and on the other hand, it can also reflect that students' preferences for curriculum types and the degree of preference are relatively stable.

Conclusion
The findings of this study showed a initial curriculum type preference scale for preschool education in vocational universities, consisting of three constructs (theory curriculum, application curriculum, and internship curriculum) and a total of 21 items.Moreover, the initial scale was tested for validity and reliability, and a final scale consisting of 20 items was developed and determined.
This study has both practical and theoretical implications.A curriculum type preference scale for preschool education in vocational universities was systematically developed for the first time, which can be used as an instrument to measure students' preferences for curriculum types in practice.In addition, the development process and methods used in this study can provide theoretical support for scale developers.
However, scale development is a long-term and complex process, and inevitably has limitations.This study used inductive methods during the scale generation phase to understand the actual status of curriculum design and serve as a reference for determining construct and developing items.However, it can hardly provide a more comprehensive understanding of the actual situation of curriculum design only through focus groups.For example, in AC1 (Implementation of Kindergarten Education Activities), due to the failure to comprehensively investigate the diverse understanding of the course type among students in reality and the failure to remove the item, the correlation between the item and the corresponding construct was weak.It can be said that there is a certain limitation in the research method used in this study.
Therefore, this study suggests that in future research, the inductive methods in the item generation phase can specifically use the questionnaire survey method, with the aim of expanding the survey scope and providing more comprehensive reference information for developing items.

Table 1
Theme 2: Application Curriculum Six experts unanimously believed that courses with both theoretical and practical elements can be classified as application curriculum, which are equipped with specialized training rooms for students to simulate and practice under the guidance of theory.
"The Montessori Teaching Method belongs to the application curriculum, as teachers need to first provide students with theoretical knowledge and then guide them to use Montessori materials in the training room" (Expert 3).Theme 3: Internship Curriculum Six experts unanimously classified pure practical courses that directly go to kindergarten to observe or participate in educational activities as internship curriculum."The courses that require students to directly enter kindergarten to observe children's behavior, or organize children's plays and teaching, are named internship curriculum by us" (Expert 4).

Table 4
The Result of Factor Retention

Table 6
The Result of Model Fit Indices

Table 7
The Result of Internal Consistency

Table 8
The Result of Test-Retest Reliability TC6 I like or good at Professional ethics and law of Kindergarten Teachers TC7 I like or good at Philosophy of