Performance Management Decision-Making Model: Case Study on Foreign Language Learning Curriculums

Abstract

Foreign language learning courses can be regarded as a service operation system, and a complete foreign language learning course performance evaluation model can help improve the effectiveness of student learning. The performance evaluation matrix (PEM) is an excellent tool for evaluation and resource management decision making, and the administrator uses the satisfaction and the importance indices to establish evaluation coordinate points based on the rules of statistical testing. The coordinate points of all service items are plotted into the PEM to grasp the full picture and to make decisions on what to improve or to consider resource transfers so as to elevate the overall satisfaction of the entire service. However, plotting all the coordinate points on the PEM can only be performed by programming, which will lead to limitations in practice. Therefore, instead of the above evaluation rules, this article uses the confidence intervals of decision-making indicators to form a validity evaluation table, to decide which teaching service items should be improved, maintained, or transferred to improve the satisfaction of the entire service system. This form of performance evaluation can be completed using any commonly used word-processing software, so it is easy to apply and promote. Finally, this article provides an applied example to illustrate the proposed method.

1. Introduction

Several studies have highlighted that globalization has influenced the demand for talent, making foreign language proficiency a critical criterion for recruitment in Taiwan’s business circles. Therefore, foreign language learning is increasingly important [1,2,3]. In the meantime, proficiency in English is also considered a fundamental requirement for admission to well-known universities worldwide. In Taiwan’s university entrance examination, English proficiency is one of the key evaluation indicators. Besides English, promoting the learning of a second foreign language has become an important policy for the universities to help students develop a second foreign language skill. Clearly, an effective foreign language course plays a crucial role in developing students’ language skills. This enhances not only their learning effect on other academic subjects but also their competitiveness for employment opportunities or further studies. In view of this, a well-rounded decision-making model of performance management for foreign language learning courses was developed to help improve students’ foreign language learning performance. Therefore, addressing this matter is of utmost importance.

Many papers are devoted to the study of the performance evaluation matrix (PEM), aiming to assess the performance of various service operations as well as to determine whether they can satisfy the requirements. Moreover, PEM is also used to decide whether to carry out improvements, maintenance, or resource transfers for the satisfaction of service items [4,5,6,7]. PEM is a great tool for evaluating and improving service-operating systems [8,9]. This approach, which focuses on the service items provided by the service-operating systems and then designs them into questionnaire scales, can be used to not only investigate customers’ or users’ satisfaction and importance for each service item but also establishes satisfaction and importance indices. Subsequently, according to the gap between these two indices for each service-operation item, three service quality zones are divided as follows: the service quality-improvement zone, the service quality-maintenance zone, and the resource-transfer zone [10]. Also, several studies have utilized confidence intervals to generate evaluation coordinate points for the satisfaction index and the importance index of each service item. This allows for an observation of the service quality zones within the PEM framework where the evaluation coordinate points for all service items lie. Consequently, decisions can be made regarding which service items require improvements or resource transfers [11].

The foreign language learning course can also be seen as a service-operating system. As mentioned above, developing a sound model of performance evaluation and resource management decision making for foreign language learning courses is an essential issue. Additionally, PEM is a great tool that is applicable to evaluation and resource management. Hence, this paper designs a teaching questionnaire using the service items which the foreign language learning courses provide to investigate students’ (customers’) satisfaction and importance for each teaching service item. According to many scholars’ viewpoints, indices are simple and unitless management tools, and they are convenient for management and communication [12]. Consequently, this paper defines the importance index, the satisfaction index, and the decision index. Nevertheless, plotting all the evaluation coordinate points into the PEM cannot be completed without programming, resulting in limitations in real practice. Therefore, this paper employs the confidence intervals of decision indices to establish a performance evaluation table (PET) to replace the PEM. This PET is built on statistical testing rules of the confidence intervals of decision indices. Accordingly, it can be used to decide on which teaching service items we should carry out improvements, maintenance, or resource transfers on to lift the satisfaction of the entire service system. To increase the improvement efficiency, the values of the decision indices can be directly used to determine the priority for making improvements when multiple teaching service items need enhancements with limited resources. As mentioned above, this approach uses the PET to solve this research gap and has advantages. In addition, regarding the questionnaire design of teaching service items, this paper adopts Yu et al.’s questionnaire involving the foreign language teaching satisfaction [13], with reference to the research of Yeh et al. [14].

The remaining sections are structured as follows. In the next section, we provide a literature review on related research studies. In Section 3, we propose research methods, including the establishment of indices involving satisfaction, importance, and decision, and then we derive confidence intervals of the decision index. Next, statistical testing is performed based on the confidence intervals. Afterward, a PET is built based on the statistical testing rules. Therefore, we can determine whether all teaching service items should conduct improvements, maintenance, or resource transfers to raise the satisfaction of the entire service system. Subsequently, in Section 4, we present an applied example illustrating the applicability of the proposed approach. Lastly, conclusions are provided in Section 5.

2. Literature Review

As mentioned above, the theme of this paper is the application of the PEM to a model of performance evaluation and resource management decision making for foreign language learning courses. This model is built on the service items that foreign language learning courses provide and later designs them into a questionnaire to investigate students’ (customers’) satisfaction and importance for each teaching service item. The research methods include setting the satisfaction index and the importance index, establishing a PEM using these two indices, and defining three service quality zones. Additionally, the decision index is defined, and the confidence interval of the index is derived. Then, the confidence interval is employed to carry out statistical testing, in order to decide whether all teaching service items should conduct improvements, maintenance, or resource transfers. When multiple teaching service items require improvements with limited resources, the decision index can directly determine the priority of improvement. As the value of the decision index is negative, the satisfaction is lower than the importance. Consequently, the priority of its improvement should be higher. Based on the above, this paper carries out a literature review on the questionnaire scales, the setting of the performance evaluation matrices and indicators, and statistical testing.

2.1. Questionnaire and PEM

Lambert and Sharma [15] built a PEM based on the seven scales of the designed questionnaire, setting satisfaction as the x-coordinate and importance as the y-coordinate, respectively. Among the seven scales, satisfaction and importance are divided into three equal parts using scales 1, 3, 5, and 7, respectively, and the PEM is divided into 9 squares. The three squares in the upper-left corner have higher importance than satisfaction. The items falling in these three squares indicate that customer importance is high, whereas satisfaction is low, so they must be improved. For the items situated in the three diagonal squares in the middle, since customers attach equal importance and satisfaction to them, they only need to maintain the current level of satisfaction. The three squares in the lower-right corner indicate that customer importance is not high, but satisfaction is high for the items. Therefore, it is necessary to examine the items and see whether they need to transfer resources to lift the overall satisfaction. Huang et al. [16] assumed that the distributions of customer satisfaction and customer importance were both subject to the BETA distribution, proposed that the index values of satisfaction and importance are between zero and one, and modified the above 9-square indices to perfect them. Then, Yu et al. [17] came up with a PEM with fuzzy semantics, encouraged the continuous improvement of comprehensive quality management to define the service quality zones, and made evaluation criteria at the same time. Many scholars have invested in research on the performance evaluation matrix to make it more complete [18,19,20]. Furthermore, several studies have applied the method of PEM to assess operational performance for various service-operating systems, aiming to enhance their overall performance. These service-operating systems include satisfaction with banking services [4], satisfaction with the service operations of various APPs [11], etc.

2.2. Index Setting and Statistical Testing

According to numerous studies, indices are standardized and unitless [12]. Whether indices are good or bad can be directly evaluated by their values without using tolerances or specifications. Therefore, they are simple and convenient management tools [21,22]. Since indices have unknown parameters, misjudgments may be caused by sampling errors, if the indices are directly evaluated by their point estimates [23]. Quite a few studies have adopted sample data to derive confidence intervals of indices and have used confidence intervals of indices to establish statistical testing rules for service operation performance indices [24,25,26]. Additionally, some studies have also employed confidence intervals of indices to set up the evaluation coordinate points of the satisfaction index and the importance index for each service item and have observed the positions of the evaluation coordinate points for all service items in the PEM, aiming to decide which service items need to perform improvements or resource transfers [10]. When the Internet of Things (IoT) becomes more and more popular and big data analysis technology matures, enterprises pursue quick responses to better their decision-making performance [27,28].

3. Research Methods

As mentioned earlier, this article uses the service items provided by foreign language learning courses to design a teaching questionnaire scale to investigate students’ satisfaction and importance of each teaching service item. An additional goal of this study is to establish a foreign language teaching performance evaluation scale based on the trust intervals of all important decision-making indicators of services and to decide which teaching service items need to be improved, maintained, or resource transferred to improve the satisfaction of the entire service system. The questionnaire content of this paper would reflect the services provided by foreign language teaching with the following five dimensions: 1. teaching preparation, 2. attitude towards teaching, 3. teaching capability, 4. teaching management, and 5. coursework and evaluation. Each dimension contains several service items that need to be investigated. Aiming to retain generality, we assumed that there were q service items that should be investigated, and the r-scale was used to conduct the survey. Moreover, each service item included two questions involving importance and satisfaction, so a total of 2q questions should be answered in the entire questionnaire [11,17]. Similar to some other papers, this paper used a random variable $Y_j$ to represent the importance of a teaching service item j and a random variable $X_j$ to represent the satisfaction of a service item j [17]. Let ${\mu }_{Yj}$ be the expected value of the random variable $Y_j$ and ${\mu }_{Xj}$ be the expected value of the random variable $X_j$. Then, we define the importance index $I_j$ and the satisfaction index $S_j$, respectively, as follows:

$$\mathrm{Importance} \mathrm{index} I_j={\displaystyle \frac{{\mu }_{Yj}-1}{R}}$$

(1)

and

$$\mathrm{Satisfaction} \mathrm{index} S_j={\displaystyle \frac{{\mu }_{Xj}-1}{R}}$$

(2)

where R $=$ k − 1. Obviously, 0 $\le$ $I_j$, $S_j$ $\le$ 1, and j = 1,…, q. According to some studies, this paper sets the satisfaction index $S_j$ as the x-coordinate and the importance index as the y-coordinate to build a PEM for foreign language teaching as follows. At the same time, the PEM was divided into three service quality zones: the service quality-improvement zone $Z_I$, the service quality-maintenance zone $Z_M$, and the resource-transfer zone $Z_T$, defined as follows:

$$\mathrm{Zone} Z_I=\left\{\left.\left(S_j,I_j\right)\right|0.18\le I_j-S_j,0\le S_j,I_j\le 1\right\}$$

$$\mathrm{Zone} Z_M=\left\{\left.\left(S_j,I_j\right)\right|-0.18<I_j-S_j<0.18,0\le S_j,I_j\le 1\right\}$$

$$\mathrm{Zone} Z_T=\left\{\left.\left(S_j,I_j\right)\right|I_j-S_j\le -0.18,0\le S_j,I_j\le 1\right\}$$

In the performance evaluation matrix for foreign language teaching, Zone $Z_I$, Zone $Z_M$, and Zone $Z_T$ are displayed in Figure 1, as follows:

Let the random variable $D_j$ $=$ $Y_j$ $-$ $X_j$ be a decision variable for the teaching service item j. Then, the expected value of the random variable $D_j$ is denoted by ${\mu }_{Dj}={\mu }_{Yj}-{\mu }_{Xj}$. Obviously, when $Y_j$ $=$ $X_j$, then the expected value of the random variable $D_j$ is ${\mu }_{Dj}={\mu }_{Yj}-{\mu }_{Xj}=$ 0. Thus, the importance of the teaching service item j is equivalent to the satisfaction. Based on Hung et al. [16], the satisfaction of the teaching service item j just needs to be maintained. When $Y_j$ $>$ $X_j$, then the expected value of the random variable $D_j$ is denoted by ${\mu }_{Dj}={\mu }_{Yj}-{\mu }_{Xj}$ $>$ 0. At this moment, the importance of the teaching service item j is higher than its satisfaction. Hung et al. [16] noted that the teaching service item j needs to be improved, as the importance of the teaching service item j is higher than its satisfaction. Similarly, when $Y_j$ $<$ $X_j$, then, the expected value of the random variable $D_j$ represents ${\mu }_{Dj}={\mu }_{Yj}-{\mu }_{Xj}<0$. At this time, the importance of the teaching service item j is lower than its satisfaction, indicating that the customer is highly satisfied with the teaching service item j but is not that attentive to it. Therefore, we only need to carry out maintenance or resource transfers to level up the satisfaction of the entire service system. Based on the above, the decision index ${\theta }_j$ is defined as follows:

$${\theta }_j={\displaystyle \frac{{\mu }_{Dj}}{R}}={\displaystyle \frac{{\mu }_{Yj}-{\mu }_{Xj}}{R}}=I_j-S_j$$

(3)

Obviously, when the value of the decision index ${\theta }_j$ is between $-$0.18 and 0.18, and $Y_j$ $=$ $X_j$, then we have $-$0.18 $\le I_j-S_j\le$ 0.18. Thus, the coordinate point of the teaching service item j belongs to $Z_M$, that is $\left(I_j,S_j\right)\in Z_M$. When the value of the decision index ${\theta }_j$ is bigger than 0.18, then we have $I_j-S_j>$ 0.18. Thus, the coordinate point of the teaching service item j belongs to $Z_I$, that is $\left(I_j,S_j\right)\in Z_I$. When the value of the decision index ${\theta }_j$ is smaller than $-$0.18, then we have $I_j-S_j<$ −0.18. Thus, the coordinate point of the teaching service item j belongs to $Z_T$, that is $\left(I_j,S_j\right)\in Z_T$. To sum up, this paper can use the value of the decision index ${\theta }_j$ to replace the PEM as the decision-making model for evaluating and improving the teaching performance of foreign language learning courses. To reduce the possibility of misjudgment due to sampling errors, we employed statistical testing rules to establish evaluation rules. The hypotheses of statistical testing for the decision index ${\theta }_j$ are listed below:

null hypothesis $H_0$: −0.18 $\le {\theta }_j\le$ 0.18

alternative hypothesis $H_1$: ${\theta }_j<$ −0.18 or ${\theta }_j>$ 0.18

Based on some other studies, the questionnaire surveys were all conducted using random sampling. Let $\left(X_{j1},\dots ,X_{jl},\dots ,X_{jn}\right)$ be a group of the random samples of teaching satisfaction and $\left(Y_{j1},\dots ,Y_{jl},\dots ,Y_{jn}\right)$ be a group of the random samples of teaching importance. Then, let $\left(D_{j1},\dots ,D_{jl},\dots ,D_{jn}\right)$ be a group of the random samples of the random variable $D_j$, where $D_{jl}=Y_{jl}-X_{jl}$, l = 1, 2, …, n, and j = 1, 2, …, q. Thus, the sample mean and sample standard deviation are expressed as follows:

$${\overline{D}}_j={\displaystyle \frac{1}{n}}\times {\displaystyle \sum _{l=1}^n{D}_{jl}}$$

(4)

and

$$S_{Dj}=\sqrt{{\displaystyle \frac{1}{n-1}}\times {\displaystyle \sum _{l=1}^n{\left(D_{jl}-{\overline{D}}_j\right)}^2}}$$

(5)

Thus, the estimator of the decision index ${\theta }_j$ is denoted as follows:

$${\widehat{\theta }}_j={\displaystyle \frac{{\overline{D}}_j}{R}}$$

(6)

The expected value and the standard deviation for the estimator of the decision index ${\theta }_j$ are defined as follows:

$$E\left[{\widehat{\theta }}_j\right]={\displaystyle \frac{E\left[{\overline{D}}_j\right]}{R}}={\displaystyle \frac{{\mu }_{Dj}}{R}}={\theta }_j$$

(7)

$$SD\left[{\widehat{\theta }}_j\right]=\sqrt{Var\left[{\widehat{\theta }}_j\right]}=\sqrt{{\displaystyle \frac{Var\left[{\overline{D}}_j\right]}{R^2}}}={\displaystyle \frac{{\sigma }_{Dj}}{\sqrt{n}R}}$$

(8)

where ${\sigma }_{Dj}$ is the standard deviation of the random variable $D_j$. Also, the random variable $Z_j$ is expressed below:

$$Z_j={\displaystyle \frac{{\widehat{\theta }}_j-{\theta }_j}{S_{Dj}/\sqrt{n}R}}$$

(9)

Based on the Central Limits Theorem, the distribution of the random variable $Z_j$ approximates the standard normal distribution, denoted by $Z_j\approx N\left(0,1\right)$ [26]. Adhering to the above principle, this paper derived the 100$\left(1-\alpha \right)$% confidence interval of the decision index ${\theta }_j$ below:

$$\begin{array}{ll}1-\alpha & =p\left(-z_{\alpha /2}\le Z_j\le z_{\alpha /2}\right)\\ & =p\left(-z_{\alpha /2}\le {\displaystyle \frac{{\widehat{\theta }}_j-{\theta }_j}{S_{Dj}/\sqrt{n}R}}\le z_{\alpha /2}\right)\\ & =p\left({\widehat{\theta }}_j-z_{\alpha /2}{\displaystyle \frac{S_{Dj}}{\sqrt{n}R}}\le {\theta }_j\le {\widehat{\theta }}_j+z_{\alpha /2}{\displaystyle \frac{S_{Dj}}{\sqrt{n}R}}\right)\end{array}$$

(10)

where $z_{\alpha }$ denotes the upper $\alpha$ quantile of the standard normal distribution. Therefore, $\left[L{\theta }_j, U{\theta }_j\right]$ represents the 100$\left(1-\alpha \right)$% confidence interval of the decision index ${\theta }_j$ as follows:

$$L{\theta }_j={\widehat{\theta }}_j-z_{\alpha /2}{\displaystyle \frac{S_{\theta j}}{\sqrt{n}}}$$

(11)

$$U{\theta }_j={\widehat{\theta }}_j+z_{\alpha /2}{\displaystyle \frac{S_{\theta j}}{\sqrt{n}}}$$

(12)

where $S_{\theta j}$ $=S_{Dj}/R$. Then, we adopted the 100($1-\alpha$)% confidence interval of the decision index ${\theta }_j$ to establish the testing rules as follows:

(1)

If $\left[L{\theta }_j, U{\theta }_j\right]\cap \left[-0.18,0.18\right]\ne \varphi$, then do not reject $H_0$ and conclude that $\left(I_j,S_j\right)\in Z_M$. Thus, the satisfaction of the teaching service item j does not require any improvement.

(2)

If $L{\theta }_j>$ 0.18, then reject $H_0$ and conclude that $\left(I_j,S_j\right)\in Z_I$. Thus, the satisfaction of the teaching service item j needs to make some improvement.

(3)

If $U{\theta }_j$ $<$ $-$0.18, then reject $H_0$ and conclude that $\left(I_j,S_j\right)\in$ $Z_T$. Thus, the satisfaction of the teaching service item j only needs maintenance or resource transfers to boost the satisfaction of the entire service system.

Following the above statistical testing decision-making rules, a foreign language teaching PET is built, as shown in

Table 1. Foreign language teaching PET.

Dimension	Item j	${\widehat{\mathit{\theta }}}_{\mathit{j}}$	$\mathit{L}{\mathit{\theta }}_{\mathit{j}}$	$\mathit{U}{\mathit{\theta }}_{\mathit{j}}$	Zone
Dimension 1: teaching preparation	1	${\widehat{\theta }}_1$	$L{\theta }_1$	$U{\theta }_1$	Z1
	2	${\widehat{\theta }}_2$	$L{\theta }_2$	$U{\theta }_2$	Z2
	⁝	⁝	⁝	⁝	⁝
	$q_1$	${\widehat{\theta }}_{q_1}$	$L{\theta }_{q_1}$	$U{\theta }_{q_1}$	$Z_{q_1}$
⁝	⁝	⁝	⁝	⁝	⁝
⁝	j	${\widehat{\theta }}_j$	$L{\theta }_j$	$U{\theta }_j$	Zj
⁝	⁝	⁝	⁝	⁝	⁝
Dimension 5: coursework and evaluation	⁝	⁝	⁝	⁝	⁝
	$q_5$	${\widehat{\theta }}_{q_5}$	$L{\theta }_{q_5}$	$U{\theta }_{q_5}$	$Z_{q_5}$

:

In Table 1, we have $q_1+q_2+q_3+q_4+q_5=q$, and then, the decision-making rules are made as follows:

(1)

When $Z_j=Z_I$, then the satisfaction of the teaching service item j requires improvement.

(2)

When $Z_j=Z_M$, then the satisfaction of the teaching service item j does not require any improvement.

(3)

When $Z_j=Z_T$, then the satisfaction of the teaching service item j only requires maintenance or a resource transfer to satisfy the entire service system.

4. An Applied Example

As mentioned earlier, we utilized Yu et al.’s foreign language teaching satisfaction questionnaire [13], which comprises five dimensions reflecting the services offered by foreign language teaching courses. Dimension 1, teaching preparation, contains 4 items, such as “The quantity of the material is appropriate”. Dimension 2, attitude towards teaching, includes 5 items, such as “Teachers value the opinions of students”. Dimension 3, teaching capability, has 2 items, such as “Teachers speak foreign languages clearly”. Dimension 4, teaching management, contains 3 items, such as “Teachers use a variety of teaching methods”. Finally, dimension 5, coursework and evaluation, contains 2 items, such as “Evaluation is at an adequate level of complexity”. These five dimensions comprise 16 question items in total, and each has two sub-questions involving importance and satisfaction. Accordingly, a total of 2q questions in the entire questionnaire need answers.

As mentioned above, the operating performance of foreign language teaching courses is the basis for students who intend to level up their foreign language proficiency. Furthermore, foreign language proficiency is often one of the key indices for recruitment opportunities in the business world. Not only can it help students improve their learning efficiency in other academic subjects, but it can also enhance their competitiveness for further studies or employment opportunities. Accordingly, foreign language teaching courses are prioritized in teaching-enhancement plans by numerous universities. Regarding the above-mentioned foreign language teaching satisfaction questionnaire consisting of 16 questions, students in the case-study school were given a total of 415 copies and were requested to return them, of which 362 copies were collected, yielding a response rate of 87.2%.

Then, the values of the relevant sample statistics for all the service items were obtained, as displayed in

Table 2. Foreign language teaching performance evaluation table with q = 16 and n = 362.

Dimension	Item	${\widehat{\mathit{\theta }}}_{\mathit{j}}$	$\mathit{L}{\mathit{\theta }}_{\mathit{j}}$	$\mathit{U}{\mathit{\theta }}_{\mathit{j}}$	Zone
Dimension 1	1	0.18	0.16	0.20	$Z_M$
	2	0.06	0.04	0.08	$Z_M$
	3	0.12	0.10	0.14	$Z_M$
	4	0.19	0.18	0.20	$Z_M$
Dimension 2	5	0.09	0.07	0.11	$Z_M$
	6	0.37	0.35	0.39	$Z_I$
	7	−0.38	−0.39	−0.37	$Z_T$
	8	0.23	0.21	0.25	$Z_I$
	9	−0.01	−0.03	0.01	$Z_M$
Dimension 3	10	0.02	0.00	0.04	$Z_M$
	11	0.05	0.03	0.07	$Z_M$
Dimension 4	12	−0.45	−0.46	−0.44	$Z_T$
	13	0.09	0.07	0.11	$Z_M$
	14	0.21	0.19	0.23	$Z_I$
Dimension 5	15	0.01	−0.01	0.03	$Z_M$
	16	−0.02	−0.04	0.00	$Z_M$

:

According to the above foreign language teaching PET, the teaching service items that need improvements include “Teachers value students’ opinions” (item 6), “Teachers are pleased to help students solve problems” (item 8) and “Teachers adequately control the pace of learning” (item 14). The teaching service items that need to take maintenance or resource transfers into account include the following: “Student-teacher interaction in class is strong” (item 7) and “Teachers use a variety of teaching methods” (item 12).

Based on the above cases, the following can be stated of the PET mentioned in this article:

(1)

The PET retains the advantages of the PEM being simple and easy to apply, and decisions can be made by the notes in the table. At the same time, you do not have to worry about the complexities and limitations of the PEM that must be drawn with the software.

(2)

The PET still conducts statistical testing based on confidence intervals to reduce the risk of misjudgments due to sampling errors.

(3)

In addition to being used to evaluate the performance of teaching systems, this model can also be applied to various service-operating systems, such as bus APPs, shopping website APPs, etc.

5. Conclusions

Enhancing the foreign language learning performance will contribute to boosting students’ foreign language proficiency. Not only can it help students level up the effectiveness of learning other academic subjects, but it can also raise their competitiveness for further studies or employment opportunities. Although the PEM can simultaneously gauge all of the service items for foreign language learning courses, its disadvantage is that the evaluation requires the support of programming software. Consequently, this paper defined the decision index based on the gap between the importance and the satisfaction indices for each teaching service item. Next, we derived a confidence interval of the decision index and performed statistical testing based on this confidence interval. This statistical testing rule retains the spirit and advantage of the decision-making rule by evaluating the coordinate points falling into the service quality zone and facilitates the decision making of administrators on whether to carry out improvements, maintenance, or resource transfers for the satisfaction of each teaching service item. Considering the confidence interval of the decision index required by the above-mentioned statistical testing rule, this paper developed a foreign language teaching PET to substitute for the foreign language teaching PEM. In fact, a performance evaluation table can be completed through word-processing software only, which is very conducive to applications and promotions. In addition, in order to boost the improvement efficiency, or when resources are limited, multiple teaching service items that need enhancements with limited resources can directly prioritize improvements based on the value of the decision index. Moreover, regarding the questionnaire design of the teaching service items, the content of the foreign language teaching course satisfaction questionnaire in this paper was divided into five dimensions. These five dimensions were employed to evaluate the services offered by foreign language learning courses. The results demonstrate that the 6th, 8th, and 14th teaching service items of dimension 2 and dimension 4 fall into $Z_I$, indicating that they need to be listed in the improvement items. Finally, this paper presented an applied example illustrating the applicability of the proposed approach.

The questionnaire in this article is designed as a K scale, which follows a multinomial distribution. It is relatively difficult and complex to make statistical inferences. A relatively large sample size is necessary to make statistical inferences by the central set limits theorem. Therefore, this study is limited to a large sample size. However, when companies pursue rapid responses and cost considerations, the sample size is usually not too large to make decisions. Future research is, therefore, recommended to consider how to develop appropriate decisions with a small sample size.