Construction of an applied model for the development of geospatial perception skills in early childhood science education

Young children’s spatial perception refers to their ability to perceive and recognize the surrounding spatial environment, including the sense of direction, distance, location and shape [1-2]. This ability plays an important role in promoting children’s physical development, motor coordination, cognitive development, and intelligence. Therefore, improving young children’s spatial perception is an important concern for educators, parents and kindergarten builders [3-6].

In order to help young children to improve their sense of direction, the following aspects can be cultivated. First of all, create a “sense of direction education environment”. Signs can be hung indoors and outdoors to indicate the four directions: east, south, west and north, so that children can gradually familiarize themselves with the concept of direction and its representation during play and learning. At the same time, maps can be made with materials of different colors or textures to guide children to observe, distinguish and express directions [7-10]. Secondly, “orientation games” can be played. In kindergarten or at home, some direction games can be played, such as “Finding Direction” game, so that children can find the designated direction according to the instructions. Or “walk the maze” game can be played to exercise children’s sense of direction and spatial reasoning ability. In addition, we emphasize the guidance in daily life [11-14]. In daily life, we can discuss the direction of walking with children, guide them to observe the surrounding environment, and perceive the change of direction while walking. At the same time, we can point to different directions with our fingers and let children imitate the movement of directions to enhance their understanding and perception of directions [15-18]. However, the above methods are not the most perfect spatial perception cultivation system, and it is important to construct a cultivation application model to enhance the spatial perception cultivation of children.

The study begins with an overview of cognitive diagnostics, focusing on cognitive diagnostics and cognitive attributes, EM parameter estimation methods, and EM algorithms for parameter estimation on the DINA and GDINA models under various combinations of conditions. Then, based on the Q-matrix theory, a strategy was developed to prepare the test questions. A small-scale pre-test was conducted, and the logical reasonableness of the test questions was tested in various aspects to form a formal diagnostic test paper on “geospatial perception ability of young children”. Using the diagnostic scale for geospatial awareness development in early childhood as the assessment target, we observed and diagnosed the selected research cases, analyzed the overall level of ability development in the cases, and specifically analyzed the achievement of the indicators in each case. The main problems in the development of geospatial perception ability of young children are summarized. At the same time, descriptive analysis was used to measure the overall geospatial competence of young children, to compare the differences between groups of young children as well as to analyze the correlation between exploratory ability and geography achievement, and finally to propose a strategy for the cultivation of young children’s geospatial perceptual competence based on the results of the experiment.

2

Overview of cognitive diagnosis

2.1

Project Response Theory Model

Item response theory assesses a subject’s ability/potential traits and item parameters by calculating mathematical models that anticipate the actual results of a subject’s responses to different test items. The model that describes the relationship between subjects’ abilities, item parameters, and response results is called the item characteristic function (ICF), and its graphical representation is called the item characteristic curve (ICC) [19]. The ability parameters and item parameters in item response theory have properties that are not present in classical measurement theory: parameter invariance and a unified scale for the ability and item difficulty parameters.

2.1.1

Rasch model

The Rasch model is a model of test analysis methodology capable of simultaneously estimating item parameters and ability parameters. The functional expression of the Rasch model is: (1) $P (Y_{i j} = 1 | θ_{i}, b_{j}) = \frac{e^{θ_{i} - b_{j}}}{1 + e^{θ_{i} - b_{j}}}$

where θ_i denotes the ability parameter for subject i and b_i denotes the difficulty parameter on item j.

2.1.2

M1PL model

The M1PL model consists of item parameters and ability parameters, given θ_i = (θ_i1, θ_i2…, θ_iD) as a vector of ability parameters for subject i and ξ_j = (a_j1, a_j2…, a_jD, b_j) as a vector of differentiation and difficulty parameters for item j. The expression for the M1PL model is: (2) $P (Y_{i j} = 1 | θ_{i}, ξ_{j}) = \frac{1}{1 + \exp (- η_{i j})}$

where $η_{i j} = \exp (\sum_{d = 1}^{D} θ_{i d} - b_{j})$ , this linear combination can provide the same sum in various combinations of potential trait values through a compensatory effect. θ_iD is the dth potential trait of subject i. b_j is denoted as the difficulty parameter of item j.

2.2

Cognitive diagnosis

Cognitive diagnostic theory is a new type of educational measurement theory that combines psychology and educational measurement to analyze the knowledge and skills needed by subjects from the perspective of cognitive psychology, to analyze the knowledge and skills needed by subjects when answering questions, and based on these analyses to incorporate a psychometric model to obtain a diagnostic information report on the strengths and deficiencies of the subjects.

2.3

Cognitive Attributes

An attribute is a piece of knowledge or skill that a student needs in order to correctly complete a question item; each attribute is actually a trait or ability. There can be more than one attribute in a question, and the same attribute can appear in more than one question. In most cognitive diagnostic models, these attributes are considered binary discrete variables of “0” or “1” [20]. The immediate purpose of cognitive diagnostic tests is to find out what attributes are mastered and what attributes are not mastered. The state of mastery of each attribute is also the state of knowledge of the subject.

2.4

Cognitive Diagnostic Model

A cognitive diagnostic model is a latent categorical model that is used to analyze the relationship between observable data and categorizable latent variables. Such latent variables are usually binary (0/1). CDMs are also known as multicategorical latent categorical models due to the qualities of the cognitive traits measured.

2.4.1

DINA model

The DINA model is a commonly used model in cognitive diagnostic assessment. The DINA model includes several elements: an item x skill attribute correlation matrix (Q matrix), parameters that represent proficiency in the skill attribute at the subject and overall levels, and item parameters (i.e., the failure parameter and the guessing parameter). q_jk in the Q matrix indicates whether skill attribute k(1 = 1, 2, …., K) is a skill attribute. q_jk in the Q matrix indicates whether the skill attribute k(1 = 1, 2, …, K) is required in order to answer item j(1 = 1, 2, …, J) correctly (0 = not required, 1 = required). Under the DINA model, α_i = {α_ik} denotes the discrete skill attribute pattern of a student i( = 1, 2, …, I) where α_ik is an indicator of whether or not the ith student has mastered the kth attribute (0 = not mastered, 1 = mastered). Based on this parameterization, the ideal answer pattern of a subject with attribute pattern α_i is denoted η_i = {η_ij} and can be expressed as: (3) $η_{i j} = \prod_{k = 1}^{K} a_{i k}^{q_{n}}$

Item failure parameter (s_i) is defined as the probability of an incorrect answer by the examinee in the case of ideal answer mode η_i = 1, while item guessing parameter (g_j) is defined as the probability of a correct answer by the examinee in the case of ideal answer mode $η_{i_{j}} = 1$ , with the formula: (4) $s_{j} = P (Y_{i j} = 0 | η_{i j} = 1) a n d g_{j} = P (Y_{i j} = 1 | η_{i j} = 0)$

The item response function of the DINA model becomes when these parameters are combined with Eq: (5) $P (Y_{i j} = 1 | α_{i}) = {(1 - s_{j})}^{η_{i j}} g_{j}^{(1 - η_{i j})}$

2.4.2

The G-DINA model

The G-DINA model also contains the Q matrix of J × K. Instead of dividing subjects’ latent traits into two categories, the G-DINA model divides them into $2^{K_{i}}$ latent traits whose $K_{j} = \sum_{i = 1}^{K} q_{j k}$ represent the number of attributes needed to correctly answer the item j. The expression of G-DINA based on log_it links is as follows: (6) $\log i t [P (a_{i j})] = v_{j ω} + \sum_{k = 1}^{K_{j}} υ_{j k} α_{i k} + \sum_{k^{'} = k + 1}^{K_{j} - 1} v_{j k k} α_{i k} α_{i k^{'}} + ... + υ_{j 12... K_{j}} \prod_{k = 1}^{K_{j}} α_{i k}$

where v_jω is the intercept parameter for item j, which represents the logit value of the item when the examinee has not mastered all the attributes needed to correctly answer the item. v_jk is the main effect parameter for knowledge state α_k, which indicates the extent to which the candidate’s mastery of attribute k contributes to correctly answering the question, $v_{j k k^{'}}$ is the interaction effect parameter between attribute k and attribute k′ on item j, and $v_{j 12 \dots K_{j}}$ is the interaction effect parameter between all attributes measured by item j.

2.5

Q-matrix theory

Q-matrix theory can be used to identify a student’s cognitive attributes that are not directly observable and transform it into a vector of observable attributes, connecting the cognitive attributes that are not directly observable to the responses that are observable in a test, thus providing a basis for further understanding and speculation about a student’s cognitive state [21]. The Q-matrix theory consists of the following basic components: the adjacency matrix (A), the reachability matrix (R), and the ideal mastery model.

The Q-matrix theory is elaborated in the following with a concrete example. The hierarchical structure of attributes in a cognitive diagnostic test is shown in Figure 1.

Firstly, according to the hierarchical relationship structure, the adjacency matrix A matrix can be obtained, which is a matrix that represents the direct relationship between attributes, that is, the indirect relationship between attributes and the relationship between themselves cannot be described, and the existing relationship is represented by “1”, and the non-existence relationship is represented by “0”. In the figure, there is a direct relationship between A1 and A2, A1 and A4, and A2 and A3, and the adjacency A matrix can be obtained. From the A matrix, the reachable matrix R matrix can be obtained by the Warshall algorithm, which is a matrix that describes the direct, indirect, and self-related relationships between attributes. Warshall’s algorithm is: in order to obtain the R array more conveniently, the algorithm is: R=(A+I)n, and I is the identity matrix. When performing matrix product (or power) operations, if the elements in the matrix are greater than “1”, a Boolean conversion is required, that is, all elements greater than “1” are converted to “1”. With (A+I) to the power (n=1,2,3,... ) increases, and when its value is stable, the matrix R is obtained. According to this algorithm, the reachable R matrix of the hierarchical relationship of the attributes can be obtained. (7) $A = | \begin{matrix} A 1 & A 2 & A 3 & A 4 \\ A 1 & 0 & 1 & 0 & 1 \\ A 2 & 0 & 0 & 1 & 0 \\ A 3 & 0 & 0 & 0 & 0 \\ A 4 & 0 & 0 & 0 & 0 \end{matrix} | R = | \begin{matrix} A 1 & A 2 & A 3 & A 4 \\ A 1 & 1 & 1 & 1 & 1 \\ A 2 & 0 & 1 & 1 & 0 \\ A 3 & 0 & 0 & 1 & 0 \\ A 4 & 0 & 0 & 0 & 1 \end{matrix} |$

After obtaining the R matrix, the expansion algorithm can be used to find out all the possible types of mastery patterns. First of all, it should be clear that each column of the R matrix also represents a mastery mode, and the expansion algorithm is to recombine these mastery modes into a new mastery mode. The specific method of the expansion algorithm is as follows: starting from the first column of the R matrix, perform Boolean addition with the subsequent columns in order, i.e., 1+1=1, 1+0=1, 0+0=0, expand the generated new columns after the original R matrix, and then loop the step from the next column in the new array until it loops to the last column of the updated matrix [22]. It is worth noting that it is possible that the columns obtained in the loop coincide with the previous ones, so that the solved columns can be discarded directly. According to this method, we can get six mastery types as (1000), (1100), (1001), (1101), (1110), (1111), and finally add the case that all attributes are not mastered (0000) to get all the mastery pattern types, which are also called ideal mastery patterns.

The ideal mastery pattern describes all logical mastery patterns that conform to the hierarchical structure of the attributes, i.e., it is not possible to obtain the cognitive situation of students outside the ideal mastery pattern after the test analysis. (8) $Ideal mastery model = | \begin{matrix} 1 & 2 & 3 & 4 & 5 & 6 & 7 \\ A 1 & 1 & 1 & 1 & 1 & 1 & 1 & 0 \\ A 2 & 0 & 1 & 1 & 0 & 1 & 1 & 0 \\ A 3 & 0 & 0 & 1 & 0 & 0 & 1 & 0 \\ A 4 & 0 & 0 & 0 & 1 & 1 & 1 & 0 \end{matrix} |$

In the testing process, there is a kind of ideal state of the subject’s response: students answer the questions without any mistakes or guesses, which is called the “ideal response mode”, i.e., students can answer the question correctly if they have mastered the attributes examined in the question, and they can’t answer the question correctly as long as there is one attribute that they have not mastered. However, in actual tests, it is impossible for a student to answer a question without making any mistakes or guesses, so the ideal response pattern is only an idealized response pattern.

3

Development of a diagnostic questionnaire on geospatial perception for young children

3.1

Development of pre-test cognitive diagnostic questionnaires

The measurement tool will directly determine the results of the diagnosis of geospatial perception ability of young children, and the preparation of the cognitive diagnostic test paper for the pre-test of young children’s map skills needs to determine the ideal mastery model of cognitive attributes, establish the Q matrix, and clarify the specific details when preparing the pre-test test paper.

3.1.1

Determining the ideal mastery model for cognitive attributes

Firstly, the adjacency matrix of six cognitive attributes of children’s geospatial perception ability was established, that is, the hierarchical relationship between cognitive attributes was expressed in the form of matrix. The matrix element is “1” or “0”, if there is a direct relationship between cognitive attributes, the adjacency matrix of the corresponding position is marked with 1, otherwise it is marked with 0. If you want to build a mental map, you need to read the map, analyze the map, and there is a direct relationship between reading the map and analyzing the map, so remember 1, and there is no direct relationship between reading the map and building a mental map in A1, so remember 0. Table 1 shows the adjacency matrix of the hierarchical relationship between children’s geospatial perceptual ability attributes.

Table 1.

Adjacency matrix

Attribute	A1	A2	A3	A4	A5
A1	1	0	0	0	0
A2	0	0	1	0	0
A3	0	1	0	0	0
A4	0	0	0	0	1
A5	0	0	0	1	0
A6	0	0	0	0	0

By using the adjacency matrix, the reachability matrix of the attribute hierarchies of geospatial perceptual abilities of young children can be further written, i.e., the attribute hierarchies that exist in the test matrix directly, indirectly, or by themselves can be represented in the form of matrices. The matrix elements are either “1” or “0”. If any of the above attribute hierarchical relationships exist between cognitive attributes, the corresponding position on the reachability matrix is marked with a 1, otherwise it is marked with a 0. The reachability matrix of the attribute hierarchical relationships of young children’s geospatial perceptual abilities is shown in Table 2. Table 2: Reachability matrix of geospatial perceptual ability Reading maps is the basis of young children’s geospatial perceptual ability, and there are certain direct and indirect relationships with all other elements, so the elements of the reachability matrix on the first row are all 1.

Table 2.

Accessibility matrix

Attribute	A1	A2	A3	A4	A5	A6
A1	1	1	1	1	1	1
A2	0	1	0	0	0	0
A3	1	1	0	1	0	0
A4	0	0	1	0	0	0
A5	0	0	0	1	1	0
A6	0	0	0	0	0	1

Based on the reachability matrix, an expansion algorithm was used to derive the ideal mastery pattern for the map skills of the children in this quiz. Starting from the first row of the reachable matrix and then adding Boolean values to each subsequent row in order, if the newly computed column is not the same as any of the previous rows, it will be added to the current reachable matrix. If the computed new column is the same as the previous row, it will be deleted until the end of the expansion algorithm loop, a new matrix will be obtained and [000000] will be added to this matrix to get the final matrix - the ideal mastery pattern matrix, the ideal mastery pattern is shown in Table 3. The rows of the ideal mastery pattern matrix represent the mastery pattern of a particular type of student, an element of 1 on the matrix means that the student has mastered the attribute, if the element is 0 it means that the student has not mastered the attribute yet.

Table 3.

Ideal mode

	A1	A2	A3	A4	A5	A6
S1	0	0	1	0	0	0
S2	1	1	0	0	0	0
S3	1	0	0	0	0	1
S4	1	1	0	0	1	0
S5	1	0	1	0	1	0
S6	1	0	0	1	0	0
S7	1	0	0	1	0	0
S8	1	0	0	1	0	0
S9	1	1	0	1	0	0
S10	1	1	1	1	0	0
S11	1	0	0	1	1	0

Through the above calculations, 11 ideal mastery patterns are finally derived, but combined with the actual teaching and life situation, [000000] such attribute mastery patterns are not of practical significance, so it is necessary to eliminate the attribute mastery pattern in the test to get the typical assessment pattern, which is shown in Table 4. The subsequent establishment of the Q matrix, the preparation of the test paper and related research will be based on the ideal mastery model and the typical assessment model.

Table 4.

Typical assessment mode

	A1	A2	A3	A4	A5	A6
S1	1	0	0	0	0	0
S2	1	0	0	0	1	1
S3	1	0	1	0	1	0
S4	1	1	0	0	0	0
S5	1	1	0	0	0	0
S6	1	0	1	1	1	0
S7	1	0	0	1	0	0
S8	1	1	0	1	0	0
S9	1	1	1	1	0	0
S10	1	1	1	1	0	0
S11	1	1	0	1	0	0

3.1.2

Establishment of the Q-matrix

The cognitive diagnostic test of young children’s geospatial perceptual ability has a total of 10 typical assessment modes, of which 10 are examining the attributes of reading maps, 7 are filling maps, 4 are drawing maps, 7 are analyzing maps, 4 are constructing mental maps, and 1 is applying maps. It can be compared and found that the examination of the attributes of drawing maps and constructing mental maps is relatively less, and the examination of the attributes of applying maps is the least, therefore, attention needs to be paid to improve the examination of the cognitive attributes of drawing maps, constructing mental maps and applying maps in the test. The Q matrix in the quiz must firstly contain all the item assessment patterns of the reachable matrix, and the rest of the assessment patterns should belong to the typical assessment patterns. The Q matrix of the quiz is shown in Table 5. The rows of the matrix represent the cognitive attributes, the columns represent the question numbers, and an element of 1 on the matrix means that the question examined the corresponding attribute, while an element of 0 on the matrix means that the question did not examine the corresponding attribute.

Table 5.

Test Q matrix

	A1	A2	A3	A4	A5	A6
Item 1	1	0	0	1	0	0
Item 2	1	0	0	1	1	0
Item 3	1	0	1	1	1	0
Item 4	1	0	1	1	1	0
Item 5	1	0	0	1	1	1
Item 6	1	1	0	1	1	1
Item 7	1	1	0	1	0	0
Item 8	1	0	1	1	0	0
Item 9	1	1	0	1	1	1
Item 10	1	1	0	1	1	1
Item 11	1	1	0	1	1	0
Item 12	1	1	1	1	1	1

3.1.3

Preparation of pre-testing questionnaires

A test question is a unit of measurement in the testing of educational and psychological traits, in which the performance of certain psychological traits (e.g., knowledge, ability, etc.) is inferred based on the subject’s responses in a prescribed stimulus situation and response format. After selecting the test content to be examined, the questions are propounded in connection with the students’ daily life situations, and the students’ thinking mode, inquiry methods and skills in geography are tested at the level of the subject. Comparing with the test Q matrix, i.e., the ideal assessment attributes, which is categorized into four cases according to the degree of consistency, namely, very consistent, relatively consistent, basically consistent and inconsistent, the degree of consistency between the actual assessment attributes and the ideal assessment attributes is shown in Table 6. From the results, it can be seen that the topics of this pre-test paper are well agreed. Therefore, this pretest paper is able to reflect the cognitive diagnostic effect of young children’s geospatial perception ability very well.

Table 6.

The consistency of the actual appraisal attribute and the ideal appraisal attribute

Test item	Actual appraisal attribute	Ideal evaluation attribute	Degree of consistency
Item 1	A1, A4	A1, A4	Very consistent
Item 2	A1, A4, A5	A1, A4, A5	Very consistent
Item 3	A1, A3, A4, A5	A1, A3, A4, A5	Very consistent
Item 4	A1, A3, A4, A5	A1, A3, A4, A5	Very consistent
Item 5	A1, A4, A5, A6	A1, A4, A5, A6	Very consistent
Item 6	A1, A2, A4, A5, A6	A1, A2, A4, A5, A6	Very consistent
Item 7	A1, A2, A4	A1, A2, A4	Very consistent
Item 8	A1, A3, A4	A1, A3, A4	Very consistent
Item 9	A1, A2, A4, A5, A6	A1, A2, A4, A5, A6	Very consistent
Item 10	A1, A2, A4, A5, A6	A1, A2, A4, A5, A6	Very consistent
Item 11	A1, A2, A4, A5	A1, A2, A4, A5	Very consistent
Item 12	A1, A2, A3, A4, A5, A6	A1, A2, A3, A4, A5, A6	Very consistent

3.2

Implementation of small-scale pre-testing

3.2.1

Purpose of pretesting

In order to further validate the rationality and reliability of the cognitive diagnostic pretest paper for map skills in young children, this study designed a small-scale pretest, which on the one hand can reduce the waste of human and material resources in large-scale tests, and on the other hand, we hope that a small number of actual responses will be given.

3.2.2

Forecasting objects

The small-scale pre-test was conducted with a kindergarten class of Primary and Secondary students, and the scoring method used was a two-point scoring model, i.e., there were only two elements of “1” or “0” for each question, with a score of 1 for a correct answer, and a score of 0 for a wrong answer, and a total score of 12 points for the question paper. The question papers will be collected immediately after the pre-test. A total of 50 test papers were distributed and 50 were collected. The pre-test experimental data are shown in Table 7.

Table 7.

Forecast test data

	Item 1	Item 2	Item 3	Item 4	Item 5	Item 6	Item 7	Item 8	Item 9	Item 10	Item 11	Item 12
ID1	1	1	1	0	1	1	0	1	1	1	0	1
ID2	1	1	0	0	1	0	1	0	1	1	1	0
ID3	0	1	0	1	0	0	1	0	1	0	1	0
ID4	1	1	1	1	1	1	1	1	1	0	0	1
ID5	1	1	0	1	1	1	0	1	1	1	0	0
ID 6	1	0	1	1	1	0	0	1	0	0	1	1
…	…	…	…	…	…	…	…	…	…	…	…	…
ID 43	1	1	0	1	0	1	1	0	1	1	1	1
ID 44	1	0	1	1	1	1	1	1	1	1	0	0
ID 45	1	1	1	1	1	1	1	1	0	0	0	1
ID 46	1	1	1	0	1	0	0	1	0	1	0	0

3.2.3

Analysis of Predictive Test Results

By calculating the difficulty of each question in the pre-test paper, the difficulty parameters of each question in the pre-test paper are shown in Table 8. As can be seen from the table, the difficulty of all the questions in the pretest paper was within the range of 0.28-0.85, with no extreme difficult or easy questions, and it was also evident that there was a certain gradient in the questions in this pretest paper. This suggests that the overall performance of this pretest paper is good and meets the basic requirements.

Table 8.

The difficulty parameters of the test paper are predicted

Issue number	difficulty	Issue number	difficulty
Item 1	0.85	Item 7	0.35
Item 2	0.79	Item 8	0.33
Item 3	0.75	Item 9	0.31
Item 4	0.52	Item 10	0.42
Item 5	0.43	Item 11	0.55
Item 6	0.52	Item 12	0.28

The results of the differentiation analysis are shown in Table 9. It can be seen that except for question 1, the differentiation coefficients of all other questions are greater than 0.3, especially the differentiation of questions 4, 5, 6, 7, 11 and 12 is very good, so it can be considered that the differentiation of this pretest paper is very good.

Table 9.

Analysis of differentials

Issue number	Differentiating	Issue number	Differentiating
Item 1	0.22	Item 7	0.55
Item 2	0.35	Item 8	0.41
Item 3	0.32	Item 9	0.32
Item 4	0.53	Item 10	0.33
Item 5	0.51	Item 11	0.57
Item 6	0.46	Item 12	0.82

4

Results of cognitive diagnosis of young children’s geospatial perceptual abilities

4.1

Diagnosis of the current situation of the development of geospatial awareness in young children

4.1.1

Diagnostic process

Two samples were randomly selected for trial observations and the observations were tested; if the results had a high level of confidence, the research community discussed the scores and determined the final evaluation results. If the reliability of the test results was insufficient, the research community familiarized themselves with the scale again and negotiated and discussed the questions to determine consistent diagnostic criteria until the reliability passed the test level. Classroom observations were conducted on the seven lesson cases and statistical tools were used to organize the evaluation results The final results of the lesson case evaluations are summarized in Table 10 (N indicates that the indicator is not represented).

Table 10.

The final results of the evaluation of the class are summarized

Index (weight)		Evaluation sample results									equipartition
Index (weight)		T1	T2	T3	T4	T5	T6	T7	T8	T9	equipartition
A(0.36)	A(0.12)	0.17	0.14	0.13	0.09	0.2	0.08	0.18	0.06	0.26	0.15
	A(0.15)	0.22	0.25	0.16	0.28	0.29	0.25	0.22	0.06	0.08	0.20
	A(0.09)	0.21	0.24	0.09	0.13	0.11	0.1	0.17	0.27	0.29	0.18
B(0.31)	B(0.09)	0.28	0.2	0.28	0.11	0.23	0.07	0.25	0.2	0.23	0.21
	B(0.11)	0.11	0.13	0.11	0.1	0.11	0.27	0.22	0.08	0.1	0.14
	B(0.11)	0.29	0.23	0.06	0.29	0.15	0.17	0.2	0.21	0.13	0.19
C(0.33)	C(0.13)	0.19	0.17	0.06	0.22	0.21	0.23	0.22	0.14	0.14	0.18
	C(0.15)	0.08	0.15	0.12	N	0.26	N	0.11	0.06	0.12	0.13
	C(0.05)	0.16	0.07	0.09	0.2	0.17	0.27	0.06	0.07	0.2	0.14
Total score		1.71	1.58	1.1	1.42	1.73	1.44	1.63	1.15	1.55
Standard score		81.62	63.55	88.62	79.54	65.96	63.51	73.55	65.63	92.63

4.1.2

Diagnostic results

1)

Overall analysis (1)

The sample as a whole is in the upper-middle level, and the differences in the scores of the lesson examples are large

The statistics of the scores of the lesson examples are shown in Figure 2. Overall, the scores of the samples range from 61.55 to 94.36. Converting them into ability levels, lesson cases T2, T6 and T8 are at the low-order-middle level, and the remaining six are at the middle-order-higher level, which indicates that the overall level of geospatial awareness development in the classroom is at the middle-upper level. In terms of the score gap, the extreme difference in the scores of the nine arbitrarily selected examples was 30.05, with T2 scoring the lowest and T10 scoring the highest, with a large extreme difference, reflecting the large internal differences in the scores of the examples. It can be seen that although the overall level of geospatial perception ability development is high, there are still some classrooms whose teaching behaviors need to be further improved.

(2)

Poor performance in the implementation and effectiveness of geospatial awareness development

The scores of the lesson cases are shown in Table 11. According to the results of the second round of expert consultation, the mean values of the scores of indicators A, B and C in the nine lesson cases were counted, which were 24.69, 22.47 and 25.96, respectively, indicating that the performance of the implementation of geospatial perceptual competence cultivation and the effect of cultivation was underperforming. In indicator A, the class cases with lower scores than the mean score are T2, T3 and T8, in indicator B, the class cases with lower scores than the mean score are T1, T2, T3, T7 and T8, and in indicator C, the class cases with lower scores than the mean score are T2 and T6, which indicates that the overall indicator scores of the T2 class cases are all smaller and need to be improved.

2)

Specific analysis

The overall analysis is a general description of the scores and indicator levels of the lesson cases, and the following focuses on the secondary indicators of the diagnostic scale and their refinement indicators, from which detailed problems in the development of young children’s geospatial perceptual ability in the classroom are found. The specific scores of the A1 indicator are shown in Figure 3. On the whole, the scores of the A1 indicators range from 3.5 to 6.2, reflecting that the goal of some lesson examples is not comprehensively directed to the acquisition of geospatial perceptual knowledge only, and the cultivation of geospatial perceptual ideas and methods is neglected. From this, it can be concluded that the objectives of the course are accurately positioned, but the cultivation of ideas and methods is neglected.

Table 11.

Score of the class

	T1	T2	T3	T4	T5	T6	T7	T8	T9	equipartition
A	31.17	20.07	17.87	20.9	35.14	32.42	27.85	18.87	17.93	24.69
B	17.55	16.73	18.17	24.03	28.12	24.5	19.57	18.28	35.27	22.47
C	28.23	17.45	33.28	26.15	22.31	19.18	27.55	25.61	33.88	25.96

4.2

Research on the measurement of geospatial perception in young children

4.2.1

Analysis of the general level of young children’s geospatial perception skills

The average score of all the assessment samples was tallied and found to be only 48.55, indicating that the overall level of geospatial perception of young children is low. The highest score is 85 and the lowest score is 15, which is a very big difference. The number of students at different score levels was counted, and the statistical histogram of the actual test sample is shown in Fig. 4, and it was found that the sample showed a normal distribution, which indicated that the test results were normal and verified the reliability of the survey data.

The water mean and standard deviation of the first-level indicators of geospatial perception ability are shown in Table 12. In terms of the three level 1 indicators of geospatial perception ability, the level of geospatial manipulation ability is the highest, followed by geospatial perception ability, and finally geospatial expression ability. The standard deviations of the three indicators are between 0.3 and 0.6, which is less than 1, indicating that the overall situation of young children’s geospatial perceptual ability is more centralized. The indicator of geospatial expression ability has the smallest standard deviation, indicating that the level of young children’s geospatial expression ability is the most concentrated, and geospatial manipulation ability has the largest standard deviation and the most dispersed data.

Table 12.

The average and standard deviation of the geographic space imagination

Variable	Spatial perception	Geographical space operation ability	Spatial expression
Average	17.25	18.65	17.62
Standard deviation	0.395	0.552	0.332

Further exploring the levels of the secondary indicator competencies of geospatial perception ability, the average scores of the results of different competency assessments were counted, and the average values of the secondary indicator water of geospatial perception ability of young children are shown in Table 13. The highest mean values were found for spatial element perception ability and spatial element manipulation ability, which were 7.35 and 7.11, respectively, and then in descending order they were: geospatial pattern expression ability, geospatial process expression ability, geospatial process perception ability, geospatial process manipulation ability, geospatial pattern manipulation ability, geospatial element expression ability, and geospatial pattern perception ability. This result was generally the same as the prespecified result, with the highest ability to perceive geospatial elements and to manipulate geospatial elements among the youngest children.

Table 13.

The mean of the secondary index of the child’s geographical space imagination

Secondary indicator	Tertiary index	Mean value
Spatial perception	Spatial factor perception	7.35
	Spatial pattern perception	3.82
	Spatial process perception	5.69
Geographical space operation ability	Spatial factor operation ability	7.11
	Spatial pattern operation ability	5.14
	Spatial process operation ability	5.52
Spatial expression	Spatial element expression ability	4.68
	Spatial pattern expression ability	5.93
	Spatial process expression	5.72

4.2.2

Analysis of group variability in young children’s geospatial perceptual abilities

1)

Analysis of gender differences in young children’s geospatial perception ability

The differences in the level of geospatial perceptual ability of young children in different groups were statistically analyzed. The ANOVA of different genders of high school students in the first level indicators of geospatial perceptual ability is shown in Table 14. As far as gender is concerned, the ANOVA shows that boys (mean score of 51.22) are not only significantly higher than girls (mean score of 49.63) in the overall level of geospatial perceptual ability, but also the mean values of the two first-level indicators of geospatial manipulative ability and geospatial expressive ability are higher than those of girls, and the mean values obtained by girls are higher than those of boys in geospatial perceptual ability. Independent samples t-test was conducted with gender as the grouping variable. The results, as shown in 4.5, show that the significance of geospatial manipulation ability and geospatial expression ability is <0.05, indicating that there is a significant difference between boys and girls in these two areas. The significance of geospatial perception ability > 0.05 indicates that there is no significant difference between male and female students in this area.

2)

Grade-level analysis of variance of young children’s geospatial perception ability

The ANOVA of the first-level indicators of geospatial perception ability of young children in different grades is shown in Table 15. As far as the grade level is concerned, the findings show that the mean value of geospatial perceptual ability scores of primary students at the overall level is 47.13, which is significantly lower than that of intermediate students (55.32 points). Moreover, the mean values were higher than those of the small class students on all three level one indicators. Independent samples t-test was conducted with grade level as the grouping variable. The significance of geospatial manipulative ability and geospatial expressive ability is <0.05, indicating that there is a significant difference between the small and intermediate grades in these two areas. Significance > 0.05 for geospatial perceptual ability indicates that there is no significant difference between the elementary and intermediate grades in this area.

Table 14.

Variance analysis of different gender high school students

Variable	Gender	Average	Standard deviation	T	Significance
Spatial perception	Man	16.32	6.9211	-1.702	0.226
Spatial perception	Female	17.85	7.021	-1.685	0.226
Geographical space operation ability	Man	18.25	9.633	-1.022	0.032
Geographical space operation ability	Female	17.05	11.523	-1.025	0.032
Spatial expression	Man	17.63	6.686	-3.436	0.015
Spatial expression	Female	15.21	5.826	-3.422	0.015

Table 15.

The variance analysis of the first grade index of children in different grades

Variable	class	Average	Standard deviation	T	Significance
Spatial perception	Small class	15.32	7.051	-3.455	0.106
Spatial perception	Middle class	18.26	6.695	-3.415	0.106
Geographical space operation ability	Small class	16.95	10.552	-1.536	0.001
Geographical space operation ability	Middle class	18.63	9.955	-1.520	0.001
Spatial expression	Small class	14.32	5.936	-4.618	0.000
Spatial expression	Middle class	17.66	6.412	-4.659	0.000

4.2.3

Correlation analysis between young children’s geospatial perception skills and academic achievement in geography

The correlations between the first level indicators of geospatial perceptual skills of young children and academic achievement in geography are shown in Table 16 (**Correlations are significant at the 0.01 level (two-tailed). *Correlations are significant at the 0.05 level (two-tailed)). Pearson was positive for all three level 1 indicators of geospatial perceptual ability, which were 0.179, 0.031, and 0.244, respectively, indicating that geospatial perceptual ability, geospatial manipulative ability, and geospatial expressive ability were all weakly and positively correlated with academic achievement in geography. Besides, the Sig values of geospatial perception ability and geospatial expression ability are less than 0.05, which means that the levels of these two abilities are significantly correlated with the level of geospatial perception ability. The Sig value of geospatial manipulation ability is greater than 0.05, which means that there is a non-significant weak correlation between geospatial manipulation ability and academic achievement in geography.

Table 16.

The correlation between the geographic space imagination and achievement

	Spatial perception	Geographical space operation ability	Spatial expression
Pearson correlation coefficient	0.179**	0.031	0.244**
Significance (double tail)	0.006	0.568	0.000

Further exploring the correlations between secondary indicators of geospatial perception and levels of academic achievement in geography, the correlations between secondary indicators of geospatial perception and geography achievement for young children are shown in Table 17 (**Correlations are significant (two-tailed) at the 0.01 level. *Correlations are significant at the 0.05 level (two-tailed)). The Sig (significance) for Geospatial Elements Perception and Geospatial Elements Manipulation competencies is greater than 0.05, indicating that the correlation with academic achievement in geography is not significant. Other than that, Sig (significance) of all the secondary indicators of competence is less than 0.05, indicating a significant correlation with the level of academic achievement in geography.

Table 17.

The correlation between the two levels

Primary indicator	Secondary indicator	Pearson correlation coefficient	Significance (double tail)
Spatial perception	Spatial factor perception	0.016	0.826
	Spatial pattern perception	0.271**	0.000
	Spatial process perception	0.116	0.046
Geographical space operation ability	Spatial factor operation ability	0.125	0.052
	Spatial pattern operation ability	0.155*	0.012
	Spatial process operation ability	0.159	0.006
Spatial expression	Spatial element expression ability	0.289**	0.000
	Spatial pattern expression ability	0.158*	0.016
	Spatial process expression	0.172**	0.006

5

Strategies for developing geospatial awareness in young children

5.1

Map reading: reading kindergarten maps

Teachers provide children with kindergarten floor plan, first guide children to recognize what is a map, explain the role of the map, and then from the kindergarten floor plan reading, let the children try to find the kindergarten on the map, the gate, playground, garden and other usual children familiar with the geographic location. In the process of map reading, children can initially feel the information-carrying function of maps, and learn to obtain more information about the spatial location of things through map information.

5.2

Mapping: Making Mental Maps

A mental map is an internalized representation of the location and characteristics of things on the earth’s surface, and the process of forming a mental map is a process in which an individual acquires information based on geospatial perception and accumulates it continuously. The drawing of mental maps should be based on a preliminary understanding of maps as a form of geographic information expression, and should be carried out in geographic environments that are familiar to young children. The kindergarten can be used as an object for drawing mental maps, and children can draw the kindergarten buildings, playgrounds and other geographic phenomena in their personal perception. Children can also try to draw a route from home to kindergarten, marking the names of landmarks and important transportation routes along the way. In order to ensure the smooth development of the mental mapping activity, the teacher needs to lead the children to observe and investigate the kindergarten, so that the children can pay attention to the important buildings or typical things along the way from home to school. The above simple mental mapping activities can help children master the spatial meanings of front and back, up and down, left and right, improve their observation and analysis of the surrounding geography, and enhance their geospatial perception.

5.3

Map location: kindergarten treasure hunt

The Learning and Development Guidelines for Children Aged 3-6 clearly recommend playing “Instruction Treasure Hunting” with children to enrich their spatial orientation recognition experience. Teachers can make a simple map of the kindergarten in advance, select several geographic locations in the kindergarten as treasure hiding places, and ask children to hold the map and describe the route to reach the treasure hiding place. Teachers can model the route to a particular cache, paying particular attention to how many meters east (south, west, north) and then how many meters north (east, south, west) the route should be. It is also important to emphasize to children that there is more than one route to any one cache, and to ask them to find as many routes as possible. Repeatedly carrying out this kind of training can let children get in touch with the concept of relative direction at the same time, train the ability of divergent thinking, so that children can grasp the spatial significance of south-east, north-west, distance and proximity, and further improve their geospatial perception ability.

5.4

Spatial reconstruction: building the kindergarten in your mind

Spatial reconstruction is a way for young children to give full play to their geospatial perceptual ability to synthesize and reproduce the outlines, shapes, sizes, directions, patterns, structures, and relationships of geographic events in geospatial space. Teachers can allow children to build a kindergarten in their mind by using blocks and other tools and combining them with the reality of the kindergarten. Since building a kindergarten is a large amount of work, it is recommended that children be divided into 3~5 groups for discussion and division of labor. In the process of building the “kindergarten in their minds”, the teacher should first guide the children to have a comprehensive understanding of the real kindergarten’s silhouette, shape and other spatial characteristics, that is, the classroom is a four-square, both length and width, and also has a height. And also note that the roof of the pointed look, then children in the selection of blocks to build a kindergarten, they can first choose the base of the kindergarten building, and then choose to have a certain area of blocks or small blocks as the walls of the building, and finally put the pointed blocks as the roof on the blocks already built. At the same time, the teacher should also actively provide detailed instructions to the children.

6

Conclusion

Aiming at how to improve young children’s geospatial perceptual ability in early education, which is a little concern in the domestic academic circles, this paper compiles a diagnostic test for the cultivation of geospatial perceptual ability based on the guidance of Q matrix theory, analyzes the results of the measurements using the DINA model, and finally proposes a strategy for the cultivation of young children’s geospatial perceptual ability based on the results of the experiments in order to provide reference and reference to the cultivation of young children’s geospatial perceptual ability in early education. Finally, based on the experimental results, we propose a strategy for the cultivation of geospatial awareness in early childhood education.

The diagnostic results show that the scores of the classroom samples range from 61.55 to 94.36, indicating that the overall level of geospatial awareness development in the classroom is in the middle to upper level. The extreme difference in the scores of the nine arbitrarily selected lesson samples is 30.05, which means that there is a large internal variation in the scores of the lesson samples. The overall level of geospatial perception ability development is high, but the teaching behaviors in some classrooms need to be further improved.

In the analysis of the overall level of geospatial perception ability of young children, the mean value of spatial element perception ability and spatial element manipulation ability was the highest, which was 7.35 and 7.11 respectively, thus it can be concluded that geospatial element perception ability and geospatial element manipulation ability were the highest in the development of geospatial perception of young children.

This study concludes that following the developmental pattern of geospatial perception ability of young children and carefully designing geography practical activities in the early education stage in a step-by-step manner can enable young children to develop their geospatial perception ability, cultivate spatial thinking, and strengthen spatial cognition in an orderly manner by carrying out activities such as reading maps, drawing maps, locating maps, and spatial reconstruction in a relaxing and enjoyable early learning process.

Language:: English

Publication timeframe:: 1 times per year
Journal Subjects:: Life Sciences, Life Sciences, other, Mathematics, Applied Mathematics, General Mathematics, Physics, Physics, other

Journal RSS Feed

	A1	A2	A3	A4	A5	A6
S1	0	0	1	0	0	0
S2	1	1	0	0	0	0
S3	1	0	0	0	0	1
S4	1	1	0	0	1	0
S5	1	0	1	0	1	0
S6	1	0	0	1	0	0
S7	1	0	0	1	0	0
S8	1	0	0	1	0	0
S9	1	1	0	1	0	0
S10	1	1	1	1	0	0
S11	1	0	0	1	1	0

	A1	A2	A3	A4	A5	A6
S1	1	0	0	0	0	0
S2	1	0	0	0	1	1
S3	1	0	1	0	1	0
S4	1	1	0	0	0	0
S5	1	1	0	0	0	0
S6	1	0	1	1	1	0
S7	1	0	0	1	0	0
S8	1	1	0	1	0	0
S9	1	1	1	1	0	0
S10	1	1	1	1	0	0
S11	1	1	0	1	0	0

	A1	A2	A3	A4	A5	A6
S1	0	0	1	0	0	0
S2	1	1	0	0	0	0
S3	1	0	0	0	0	1
S4	1	1	0	0	1	0
S5	1	0	1	0	1	0
S6	1	0	0	1	0	0
S7	1	0	0	1	0	0
S8	1	0	0	1	0	0
S9	1	1	0	1	0	0
S10	1	1	1	1	0	0
S11	1	0	0	1	1	0

	A1	A2	A3	A4	A5	A6
S1	1	0	0	0	0	0
S2	1	0	0	0	1	1
S3	1	0	1	0	1	0
S4	1	1	0	0	0	0
S5	1	1	0	0	0	0
S6	1	0	1	1	1	0
S7	1	0	0	1	0	0
S8	1	1	0	1	0	0
S9	1	1	1	1	0	0
S10	1	1	1	1	0	0
S11	1	1	0	1	0	0

Construction of an applied model for the development of geospatial perception skills in early childhood science education

Debin Si

Published Online: Sep 26, 2025

Received: Jan 25, 2025

Accepted: Apr 30, 2025

DOI: https://doi.org/10.2478/amns-2025-1043

KeywordsQ-matrix theory, DINA model, Geospatial perception ability, Early childhood science education

© 2025 Debin Si, published by Sciendo.

This work is licensed under the Creative Commons Attribution 4.0 International License.

Keywords
Q-matrix theory, DINA model, Geospatial perception ability, Early childhood science education

	A1	A2	A3	A4	A5	A6
S1	0	0	1	0	0	0
S2	1	1	0	0	0	0
S3	1	0	0	0	0	1
S4	1	1	0	0	1	0
S5	1	0	1	0	1	0
S6	1	0	0	1	0	0
S7	1	0	0	1	0	0
S8	1	0	0	1	0	0
S9	1	1	0	1	0	0
S10	1	1	1	1	0	0
S11	1	0	0	1	1	0

	A1	A2	A3	A4	A5	A6
S1	1	0	0	0	0	0
S2	1	0	0	0	1	1
S3	1	0	1	0	1	0
S4	1	1	0	0	0	0
S5	1	1	0	0	0	0
S6	1	0	1	1	1	0
S7	1	0	0	1	0	0
S8	1	1	0	1	0	0
S9	1	1	1	1	0	0
S10	1	1	1	1	0	0
S11	1	1	0	1	0	0