JP2005215023A - Test implementation system and test implementation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a test implementation system and a test implementation method with which a capability value having a proper reliability coefficient is given in a computer adaptive test. <P>SOLUTION: The computer adaptive test implementation system 12 comprises: a capability value calculation section 20 for calculating, on the basis of whether an item is correctly answered or not, the capability value of an examinee; an item selection section 18 for selecting, on the basis of the calculated capability value, prescribed items out of an item database storing test items; and a reliability coefficient calculation section 22 for calculating the reliability coefficient of the capability value. The item database stores the items belonging to a plurality of item groups whose question setting order is prescribed. The item selection section 18, when the reliability coefficient of the capability value reaches the minimum value set for each item group, selects the item belonging to the next item group according to the question setting order. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a test execution system and a test execution program that give questions that match the ability of the examinee.

  With the spread of the Internet, the computer of the management center presents a problem to the examinee and asks for the input of the answer using an electronic terminal, and the computer of the management center automatically answers the answer composed of the input answer. A system has been proposed in which scoring is performed and the scoring results are stored in a large-scale storage device.

In such a system, test items corresponding to the ability level of the examinee are selected from the test item database storing the questions, and the examinee inputs an answer by operating the input device of the electronic terminal. It is supposed to be. In particular, the item response theory (IRT) is used to estimate the estimated value of the ability of the examinee based on the correctness of the test item, and to optimize the next question. .
JP 2001-209297 A "Statistics Library Item Response Theory [Introduction]-Science of Testing and Measurement" (Hideki Toyoda, April 15, 2002, Asakura Shoten, pages 61-75)

  In tests using the item response theory as described above, parameters such as difficulty and discriminatory power, and the estimated value of the testee's ability value based on the correctness of the test item (hereinafter referred to as ability value) have a certain level of reliability. It is desirable to reach. A reliability coefficient is used as an index representing the reliability. For this reason, in the test described above, the test items are repeated until the ability value reaches a certain level of reliability. However, there are cases where it is difficult to ask questions until a certain level of reliability is reached, due to the progress of the examinee's responses and the time limit of the test.

  Moreover, in the test using an electronic terminal, the number of questions is not determined in advance and varies depending on the progress of the examinee. Therefore, unlike the normal paper test, there is also a problem that the examinee cannot know how far he has progressed.

  An object of the present invention is to provide a test execution system and method capable of obtaining an appropriate reliability capability value. It is another object of the present invention to provide a test execution system and method that enable the examinee to grasp his / her progress during the test.

  An object of the present invention is a computer-adaptive test execution system in which items for evaluating the ability of a test subject are selectively set for each test subject, and based on the correctness of the answer of the test item, An ability value calculating means for calculating an ability value; an item selecting means for selecting a predetermined item from an item database storing test items based on the calculated ability value; and a reliability coefficient for the ability value is calculated. The item database stores items belonging to a plurality of item groups in which the order of questions is defined, and the item selection unit has a reliability coefficient of the ability value as the item group. This is achieved by a test execution system configured to select an item belonging to the next item group according to the order of questions when a minimum value set for each item is reached.

  In a preferred embodiment, when the item selection means asks a predetermined number of items or more in the certain item group, the reliability coefficient of the ability value is the reliability set for each item group. Even when the reference value (S_std) has not been reached, the next aspect (item) is selected with priority on the elapsed time.

  In another preferred embodiment, the item database stores a first item that inquires about the first aspect and a second item that inquires about the second aspect, and the item selection means includes the first item. When the reliability coefficient of the ability value related to the item group reaches a certain threshold value, the item related to the second aspect is selected. In the above embodiment, it is possible to adapt a linguistic aspect as the first aspect and a non-verbal aspect as the second aspect.

  In a more preferred embodiment, when the item selection means has passed the first time limit even if the reliability of the capability value related to the first aspect has not reached a certain threshold value, It is comprised so that the item regarding two side surfaces may be selected.

  Furthermore, in a preferred embodiment, the item selecting means refers to the reliability coefficient related to the first aspect when the reliability coefficient of the capability value related to the second aspect reaches a certain threshold value. If necessary, the first item is selected.

  In another preferred embodiment, a display field for generating an answer status display field for notifying the progress status of the test is further displayed on the display device of the terminal device for presenting the item and allowing the examiner to input an answer. The display field generation means is configured to advance a variable bar in the answer status display field as the item group progresses. A pie chart display field can be used as the answer status display field. In this case, for example, a variable bar extending in the radial direction from the center advances clockwise.

  Another object of the present invention is a computer-adaptive test execution method for selectively setting items for evaluating the ability of the examinee for each examinee, based on whether the answer of the item is correct or not. An ability value calculating step for calculating the ability value of the examinee, an item selecting step for selecting a predetermined item from an item database storing test items based on the calculated ability value, and reliability of the ability value A reliability coefficient calculation step for calculating a coefficient, and the item database stores items belonging to a plurality of item groups in which the order of questions is defined, and the item selection step includes a reliability coefficient of the ability value. And a test execution method comprising a step of selecting an item belonging to the next item group according to the order of questions when the minimum value set for each item group is reached Also it is achieved by.

  In a preferred embodiment, when the item selection step asks a predetermined number of items or more in the certain item group, the reliability coefficient of the ability value reaches a minimum value set for each item group. If not, the method includes a step of selecting an item belonging to the next item group according to the order of questions.

  In another preferred embodiment, the item database stores a first item for inquiring a first aspect and a second item for inquiring a second aspect, and the item selecting step includes the step of Selecting a second item relating to the second aspect when the reliability coefficient of the capability value relating to the first aspect reaches a certain threshold value;

  More preferably, in the case where the item selection step has passed the first time limit even if the reliability coefficient of the ability value related to the first aspect has not reached a certain threshold value, the second aspect Selecting the second item for.

  Further, the item selection step refers to the reliability coefficient related to the first aspect when the reliability coefficient of the capability value related to the second aspect reaches a certain threshold, and if necessary, the A step of selecting the first item may be included.

  In another preferred embodiment, a display field for generating an answer status display field for notifying the progress status of the test is further displayed on the display device of the terminal device for presenting the item and allowing the examiner to input an answer. A generation step, and the display column generation step includes a step of advancing a variable bar in the answer status display column as the item group progresses.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to provide the test execution system and method which can obtain the capability value of a suitable reliability coefficient. In addition, according to the present invention, it is possible to provide a test execution system and method in which the examinee can grasp his / her progress during the test.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing a configuration of a test execution system according to an embodiment of the present invention. As shown in FIG. 1, the test execution system 10 includes a test operation server 12 and a plurality of test terminals 14-1, 14-2,... Connected to the test operation server 12. The network connecting the test terminal 14 and the test operation server 12 can use a LAN or WAN.

  The examinee uses his / her user terminal (not shown) to access the test operation server 12 via a network (not shown) such as the Internet, and reserves the date and place for taking the test. The test is conducted by operating the designated test terminal at the reserved date and time and place.

  As shown in FIG. 1, the test operation server 12 includes a test item database (DB) 16 that stores various test items, an item selection unit 18 that selects test items from the test item DB 16, and an input from the examinee. Based on the determination, the ability value calculation unit 20 that calculates the ability value of the examinee, the reliability coefficient calculation unit 22 that calculates the reliability coefficient of the ability value of the examinee, the test time and the examination A progress display generating unit 24 for generating a progress display to be displayed on the screen of the display device of the test terminal according to the progress of the tester's item, and a test result for storing test results such as the correctness / incorrectness of the tester's item and ability DB26.

  Initially, as will be described later, the item selection unit 18 sequentially selects and extracts the initial items for acquiring the ability level of the examinee from the test item DB 16 and presents them to the test terminal 14. , Urge the examinee to answer. Thereafter, an item is selected from the test item DB 16 according to the ability value based on the correctness / incorrectness determination for the problem by the ability value calculation unit 20. Note that each component of the test operation server 12 is configured so that items are given according to each ability value of the examinee and the progress of the test. Therefore, even when a plurality of examinees take the test at the same time, the items to be given are different.

  Hereinafter, the question logic of the test question item in the present embodiment will be described. In the present embodiment, the test question includes an item group for asking for ability and an item group for asking for personality. Further, the item group for inquiring about the ability aspect includes an item group for looking at linguistic ability and an item group for looking at non-language ability. Hereinafter, an item group for viewing linguistic ability is also referred to as a language area, and an item group for viewing non-linguistic ability is also referred to as a non-language area.

  For the item for which the ability aspect is questioned, upon receiving the input (answer) to the test terminal 14 by the examiner, the ability value calculation unit 20 determines whether the answer is correct or not, and determines the answer and correctness of the examinee based on the test result. In DB26, it memorize | stores in the record of ID which pinpoints the said examinee. The correctness / incorrectness of the answer is transmitted to the item selection unit 18 and is used for selecting an item to be asked next, and the ability value calculated based on the correctness / incorrectness of the answer is determined by the reliability coefficient calculation unit 22. The reliability coefficient is calculated. On the other hand, since there is no right or wrong answer for the item for which the personality is questioned, the item selection unit 18 selects from the test item DB 16 in a predetermined logic (for example, random), and the answer received from the test terminal 14 Only in the record of the ID that identifies the examinee in the test result DB 26.

  FIG. 2 is a block diagram schematically showing the question logic of items for which the ability aspect is asked. As shown in FIG. 2, in the present embodiment, the item selection unit 18 first reads the language area items (initial items) from the test item DB 16 and presents them on the screen of the display device of the test terminal 14. (Step 201). The examinee operates the input device of the test terminal 14 to input an answer to the presented item, transmits it to the test operation server 12, and the test operation server 12 receives the transmitted answer. This process is repeated a predetermined number of times (see step 202). At this time, the ability value calculation unit 20 calculates an initial ability value for the linguistic ability based on the correctness of the initial item in the language area (step 203). When the initial ability value is calculated, the item selection unit 18 selects and reads out an appropriate language area item from the test item DB 16 according to the calculated ability value, and displays the display device of the test terminal 14. (Step 204). Upon receiving the examinee's answer, the ability value calculation unit 20 recalculates the ability value of the language area based on the correctness of the answer (step 205). Further, the reliability coefficient calculation unit 22 calculates a reliability coefficient of the capability value (step 206). The progress display generation unit 24 determines the position of the variable bar based on the capability value and its reliability. Such processing is repeated until a predetermined condition is satisfied (see step 207).

  For items in the non-language area, items are presented with similar logic (step 210). Then, according to the conditions described later, in order to increase the reliability of the respective ability values, questions in the language domain and / or items in the non-language domain are presented (step 211).

  In addition to the above question logic, in the present embodiment, time constraints are added to the questions of items in each area. The time limit for questions in the language area is set to a predetermined value T_1. As will be described later, when the time T_1 has elapsed, the items in the non-language area are given without any questions in the language area. Further, in the present embodiment, as described above, items in the non-language area are given after the questions in the language area. Here, a predetermined value T_3 (> T_1) is given as a time limit for questions in the non-language area. In addition, T_max (> T_3) is defined as the maximum time for questions that ask the ability aspect. Regarding time management in a non-linguistic domain, even if T_max is exceeded, the items that have been asked are allowed to answer within the time limit (for each item). Therefore, when T_2 (T_1 <T_2 <T_3) has elapsed so that the implementation time does not greatly exceed T_max, control is performed so that only the problem with a relatively short time limit is presented.

  In the following, the selection of items for which capability aspects are asked, the capability value, and the reliability thereof according to the present embodiment will be briefly described. In the present embodiment, an item reaction theory is used that calculates a probabilistic score (ability value) based on whether or not an answer to an item is correct. Furthermore, in the present embodiment, an adaptive test is adopted in which the next item to be set is selected based on the ability value.

Each item of the test has a “difficulty” that is an index indicating the difficulty of the item, and a “discriminating power” that is an index that indicates whether the measurement accuracy of the item is good or bad. For example, as shown below, an item having a combination of “difficulty” and “discrimination” is given, and the combination of “difficulty” and “discrimination” of the next item is determined based on the correctness. .
(1) “Difficulty: High” and “Distinguished: Medium” Error (2) “Difficulty: Medium” and “Distinguishing power: High” Correct answer (3) “Difficulty: High” and “Distinguishing power” : Medium "Error (4)" Difficulty: Medium "and" Distinguishing power: High "Error:
Such a process is repeated and the test proceeds. As the capability value, for example, an estimated value θ by Bayesian estimation can be used. Moreover, as shown in Non-Patent Document 1, the following equation (1) can be used as a reliability coefficient representing reliability.

上述の式において標準誤差（σ_ｅ：誤差分散の平方根）は以下の（２）式を利用して算出される。

In the above formula, the standard error (σ _e : square root of error variance) is calculated using the following formula (2).

なお、フィッシャー情報量は、２変数の場合に、以下の（３）式で表わすことができる。

The Fisher information amount can be expressed by the following equation (3) in the case of two variables.

ここで、フィッシャー情報量を用いない場合、例えばベイズ推定を用いることができる。前述したように、本実施の形態においては、能力的側面を問う項目群に、言語領域の項目群および非言語領域の項目群が含まれる。そこで、言語領域および非言語領域のそれぞれについて、能力値およびその信頼性が算出されるようになっている。言語領域の能力値の信頼度については、単一の閾値Ｓ＿ｓｔｄが設定される。非言語領域の項目群は、初期項目に属する項目群と、６つの項目群とが設けられ、６つの項目群のそれぞれについて、能力値の信頼性のジャッジポイント（Ｓ＿ｍｉｎ−ｚｈ１〜Ｓｍｉｎ−ｚｈ６）が設けられる。能力値の信頼性（信頼性係数）がジャッジポイント以上になると、次の項目群に属する項目が選択可能となる。また、言語領域と同様に、非言語領域についても、閾値Ｓ＿ｓｔｄが設定される。

Here, when the Fisher information amount is not used, for example, Bayesian estimation can be used. As described above, in the present embodiment, the item group inquiring about the capability aspect includes the item group in the language area and the item group in the non-language area. Therefore, the ability value and its reliability are calculated for each of the language area and the non-language area. A single threshold value S_std is set for the reliability of the ability value of the language domain. The item group of the non-language area includes an item group belonging to the initial item and six item groups. For each of the six item groups, a judgment point (S_min-zh1 to Smin-zh6) of the reliability of the ability value. Is provided. When the reliability (reliability coefficient) of the ability value is equal to or higher than the judgment point, items belonging to the next item group can be selected. Similarly to the language area, the threshold value S_std is set for the non-language area.

  As schematically described above, the present embodiment is configured such that the next item is selected based on the correctness of the answer for the item asking for the capability aspect. The answer time per question is limited, and if the user answers within a predetermined time and is instructed to proceed to the next problem, the management server 18 presents the next problem according to whether the answer is correct or incorrect. To do. In some cases, the user does not instruct to proceed to the next item even if the answer is made within a predetermined time. In this case, after a predetermined time elapses, the test operation server 12 determines whether the user's answer is correct, and presents the next item according to the result.

  On the other hand, if the user does not answer within the predetermined time, the next item is presented in the same manner as if the answer was incorrect, with no answer for that item. The progress of the test items varies depending on the user's response. In the test using the test terminal 14, the user may not know the total number of problems and may not be able to grasp his / her progress. Therefore, in the present embodiment, a column for displaying the overall progress and elapsed time is provided in the test question image on the display device of the test terminal 14 so that the user can grasp the progress and remaining time. It is configured. Such a display field may be displayed at the upper right corner of the image, for example.

  FIG. 3 is a diagram illustrating an example of a display field. As shown in FIG. 3A, the display column 300 is a double pie chart, and the inside shows the answer status and the outside shows the passage of time.

  The answer status display column 301 displays the progress of the answer. The time passage display field 302 has a function equivalent to that of a clock. In the present embodiment, the test items include a test item for viewing “capability aspects” and a test item for viewing “personal aspects”. The display logic of the answer status display column and the time progress display column is different for each.

First, during the test, we will explain the questions for questions about ability and the display of the answer status display column. In the answer status display column, the current progress of the examinee is represented by a variable bar (for example, reference numeral 303). The progress of the variable bar in the answer status display field follows the following logic. As the variable bar progresses, the area between the initial position and the variable bar is displayed differently from the other areas (see hatching in FIG. 3A). For example, display such as hatching or shading may be used, or display may be performed using different colors.
(1) The variable bar advances clockwise in the answer status display column. (2) In the language area, when the questions of the initial item and the prepared area are completed, the variable bar advances by a predetermined angle. Proceed when new issues in this area are presented. Here, the region includes, for example, a region for measuring vocabulary, a region for measuring ability related to language usage (grammar), and a region for measuring ability related to language similarity determination.
(3) In the non-language area, the variable bar advances by a predetermined angle every time the answer of the initial item is finished, and thereafter, every time the answer is finished, it is determined whether or not the set reliability is reached, The variable bar will advance if it has been reached, and will not advance if it has not been reached. However, if the variable bar stays at the same place, there is a possibility that it may cause anxiety to the examinee. Once the variable bar stays, if there is an answer to the next question, whether or not the above reliability has been reached. Regardless, the variable bar advances.
(4) The variable bar makes one round and returns to the initial position only when a reliable threshold value is reached for both the language problem and the non-language problem.

  The question of items for inquiring about the capability aspect and the progress of the variable bar in the answer status display column will be described with reference to FIG. 3B and FIGS. 4 to 6 which are detailed processing flows.

  Each time the item selection unit 18 sets a language item (step 401), the progress display generation unit 402 advances the variable bar 15 degrees clockwise (step 402). Here, a predetermined number of items presented at the beginning are initial items in the language area.

  In the present embodiment, the language area items are basically set to 8 areas. Therefore, the variable bar reaches the position “a” at the stage where the items of the language area are set to 8 areas. If it is determined yes in step 403, it is further determined whether a problem in the language domain can be set (step 403). In the present embodiment, eight or more language area items are prepared, and if possible, the item selection unit 18 selects an appropriate language area item from the test item DB 16 and gives a question. In this case, the variable bar is maintained at position “a” (step 405).

  On the other hand, if the number of questions in the language area is less than 8 (No in step 403), it is determined whether or not the condition for ending the questions in the language area is satisfied (step 406). .

  The termination condition in step 406 will be described below. The termination condition includes a time condition and a condition related to reliability. When the elapsed time from the start of the test exceeds T_1 for the time condition (step 406), the progress display generation unit 24 advances the variable bar to the position “a (see FIG. 3B)”. (Step 407). In other words, even if the items in the language area cannot be set as eight questions as in the basic case, the question of the items in the language area is discontinued when the time limit T_1 has elapsed, and the item selection unit 18 performs the test item DB 16. Then, the initial items in the non-language area are selected and the questions are given (step 408).

  Further, regarding the reliability condition, if the value of the language domain reliability coefficient S1 is equal to or greater than a predetermined threshold value S_std, it is determined as Yes in step 406. In other words, if the language area items cannot be set to 8 questions as in the basic case, and if the language area reliability coefficient S1 ≧ S_std, it is determined that the language area items have been sufficiently set. As a result, the progress display generation unit 24 moves the variable bar to “a”, and then the item selection unit 18 selects an initial item in the non-language area from the test item DB 16 and gives a question (step 408).

  In addition, when it is determined NO (No) in step 404, that is, when the items in the language area are exhausted, the item selection unit 18 selects the initial items in the non-language area from the test item DB 16. Select and give a question (step 408).

  When the initial items in the non-language area are presented, the progress display generator 24 advances the variable bar by 25 degrees (step 409). This process is repeated until the initial questions in the non-language area are completed (see step 410). In the present embodiment, since three items are given as initial items, the variable bar reaches the position “b” in FIG. 3B when the questions of the initial items are completed.

  If the item in the non-language area is set (Yes in Step 501), it is determined whether or not the position of the variable bar has reached “h” (Step 503). If it is determined NO in step 501, the position of the variable bar is maintained (step 502).

  As described above, in the present embodiment, in the non-language area, in addition to a predetermined number of initial items, six item groups are provided, and in each item group, the reliability coefficient S2 of the non-language area is set. Judge points Smin-zh1 to Smin-zh6 are provided. If it is determined NO in step 503, whether or not the reliability coefficient S2 is equal to or higher than the judgment point Smin-zh “n” (n = 1 to 6) corresponding to the item group. Is determined (step 504).

  If it is determined NO in step 504, it is then determined whether or not two or more items have already been set in the item group (step 505). If the number of questions is less than 2 (No in Step 505), the position of the variable bar is maintained as it is, and the process returns to Step 501, and the items in the non-language area belonging to the item group are presented ( Step 501).

  On the other hand, if the reliability coefficient S2 of the non-language area is equal to or larger than the judgment point Smin-zh “n” of the item group (Yes in Step 504), the reliability coefficient S2 of the non-language area is the threshold value S_std. It is determined whether or not the above is true (step 507). If NO is determined in step 507, the variable bar is advanced by one step (step 508), and the item in the non-language area of the next item group is selected and presented (step 501). As described above, in the present embodiment, when the reliability coefficient S2 is equal to or higher than the minimum reliability coefficient Smin-zh “n” for each item group in the non-language area, the variable bar is advanced by one step. In addition, items belonging to the next item group are presented. As a result, the reliability coefficient can be gradually increased. In the progression of the variable bar in step 508, the variable bar is advanced step by step from position “b” to “h” in FIG. If the non-language domain reliability coefficient S2 is equal to or greater than the threshold value S_std (Yes in step 507), the variable bar advances to “h” (step 509). In parallel with the processing shown in FIG. 5, it is determined whether or not the time limit T_3 for questions in the non-language area has been reached. If the non-language region reliability coefficient S2 is equal to or greater than the threshold value S_Std before the time limit T_3 is reached, the variable bar proceeds to “h” and shifts to the processing shown in FIG. However, as shown in FIG. 7, when the time T_3 has elapsed (Yes in Step 701) and the reliability coefficient S2 is not equal to or greater than the threshold value S_std, that is, the position of the variable bar is “b” to “g”. ”(No in step 702), the questions of the non-language area items are continued (step 703). Here, the ability value based on the answer to the question and its reliability coefficient are calculated, and if the reliability coefficient S2 is equal to or higher than the judgment point (step 704), the variable bar is advanced (step 705). The processes (steps 706 and 707) executed when the reliability coefficient S2 is less than the judgment point (No in step 704) correspond to steps 505 and 506 in FIG. 5, respectively.

  When the position of the variable bar reaches “h”, it means that the questions in the non-linguistic area are substantially finished. From then on, the remaining time is used to refer to the reliability coefficients of the language area and the non-language area, and in order to further improve the reliability, the items of the language area and / or the items of the non-language area are set as questions. Is done.

  As shown in FIG. 6, if the reliability coefficient S2 of the non-language area is less than the threshold value S_Std due to the questions of the language area items in accordance with the processing in FIG. 4 (No in step 601), it is variable. While maintaining the position of the bar at “h” (step 602), items in the non-linguistic area are presented, and then the ability value based on the correctness of the answer and the reliability coefficient thereof are calculated. Return.

  If the result of step 601 is yes (Yes), if the reliability coefficient S1 of the language area is less than the threshold value S_Std (No in step 604, the position of the variable bar is maintained at “h”) In step 605, it is determined whether or not the language area can be given a question (step 606) If the item in the language area can be given a question (Yes in step 606), the item selection unit 18 Then, the language area item is selected from the test item DB 16. Then, the ability value and the reliability coefficient are calculated based on the correctness of the answer, and the process returns to step 604.

  If it is determined NO in step 606, it corresponds to the case where all the questions in the language area have already been set. In this case, when the reliability coefficient S2 of the non-language region becomes equal to or greater than the threshold value S_std (Yes in Step 601) and No in Step 606, the progress display generation unit 24 sets the variable bar to the initial position (that is, The time is advanced to the 12 o'clock position of the clock (step 608). As a result, the examinee can feel that all the questions have been completed.

  Thus, according to the present embodiment, the ability value is calculated according to the answer to the question, and the reliability coefficient of the ability value is calculated. The following items are given based on the reliability coefficient and the time limit. The progress display generation unit 24 controls the progress of the variable bar in the answer status display column in accordance with the questioned items and the reliability of the ability value. Therefore, even in a test in which the number of questions is not determined in advance, it is possible to notify the examinee of the progress of the test.

  Next, the time passage display field will be described. The initial position of the variable bar in the time passage display column is also the 12 o'clock position of the clock. As shown in FIG. 8, the display related to the time passage display field is executed every minute (see step 801). The display progress generation unit 18 advances the variable bar clockwise by “120 / T_1” degrees until the elapsed time reaches the time T_1 from the start of the test (Yes in Step 802) (Step 803). That is, when the time T_1 has elapsed, the variable bar reaches a position corresponding to the position “a” in FIG. Time T_1 is a time limit for questions in the language domain. Therefore, when the answer of the item in the language area is completed according to the time limit, the positions of the variable bar in the time passage display field and the variable bar in the answer status display field coincide.

  After time T_1, the display progress generation unit 18 advances the variable bar clockwise by “235 / (T_max−T_1−1)” degrees every minute (steps 804 and 805). Thereby, when the maximum time T_max has elapsed, the variable bar reaches a position rotated by 355 degrees from the initial position. When the test is finally completed, or when the examinee operates the test terminal and gives an instruction to end the test (Yes in Step 806), the display progress generation unit 18 sets the variable bar five times. The variable bar is returned to the initial position by moving it clockwise only (step 807).

  Thus, according to the present embodiment, it is possible to appropriately notify the examinee of the progress of time while taking into account the progress of items.

  The number of items for which questions about personality are asked is basically determined. Accordingly, in the answer status display field, the variable bar may be advanced clockwise by (360 / number of items) every time there is an answer. For the time-lapse display field, it is sufficient to advance the variable bar clockwise by (360 / T_max) degrees.

  The present invention is not limited to the above embodiments, and various modifications can be made within the scope of the invention described in the claims, and these are also included in the scope of the present invention. Needless to say.

FIG. 1 is a block diagram showing a configuration of a test execution system according to an embodiment of the present invention. FIG. 2 is a block diagram schematically showing the question logic of items for which the ability aspect is asked. FIG. 3 is a diagram showing an example of display columns displayed on a part of the screen of the display device of the test terminal in the present embodiment. FIG. 4 is a flowchart showing the progress of a variable bar in an answer question display column and questions for questions about ability aspects in the present embodiment. FIG. 5 is a flowchart showing the progress of a variable bar in an answer question display column and questions for questions about ability aspects in the present embodiment. FIG. 6 is a flowchart showing the progress of a variable bar in an answer question display column and questions for questions about ability aspects in the present embodiment. FIG. 7 is a flowchart showing the progress of a variable bar in an answer question display column and questions for questions about ability aspects in the present embodiment. FIG. 8 is a flowchart showing the progress of the variable bar in the time passage display column in the present embodiment.

Explanation of symbols

10 Test Execution System 12 Test Operation Server 14 Test Terminal 16 Test Item DB
18 Item selection unit 20 Capability value calculation unit 22 Reliability coefficient calculation unit 24 Progress display generation unit 26 Test result DB

Claims (12)

A computer-adaptive test execution system in which items for evaluating the ability of a test subject are selectively given to each test subject,
Ability value calculation means for calculating the ability value of the examinee based on whether the answer of the item is correct,
Based on the calculated ability value, an item selection means for selecting a predetermined item from an item database storing test items;
A reliability coefficient calculating means for calculating a reliability coefficient of the capability value;
In the item database, items belonging to a plurality of item groups in which the order of questions is defined are stored,
The item selecting means is configured to select an item belonging to the next item group according to the order of questions when the reliability coefficient of the ability value reaches a minimum value set for each item group. A test execution system characterized by that. When the item selection means asks a predetermined number of items or more in the certain item group, even if the reliability coefficient of the ability value does not reach the minimum value set for each item group, The test execution system according to claim 1, wherein the test execution system is configured to select an item belonging to a next group of items according to an order. The item database stores a first item for inquiring about a first aspect and a second item for inquiring a second aspect,
The item selection means is configured to select the second item related to the second aspect when a reliability coefficient of the capability value related to the first aspect reaches a certain threshold value. The test execution system according to claim 1 or 2, characterized in that Even when the reliability coefficient of the capability value related to the first side group has not reached a certain threshold value, the item selection unit is configured to execute the first side time limit when the first time limit has elapsed. The test execution system according to claim 3, wherein the test execution system is configured to select a second item. The item selection means refers to the reliability coefficient related to the first aspect when the reliability coefficient of the capability value related to the second aspect reaches a certain threshold value. The test execution system according to claim 4, wherein the system is configured to select one item. In addition, the display device of the terminal device for presenting the item and allowing the examiner to input an answer is provided with a display field generating means for generating an answer status display field for notifying the progress of the test,
The test execution system according to claim 1, wherein the display field generation unit is configured to advance a variable bar in the answer status display field as the item group progresses. A computer-adaptive test execution method in which items for evaluating a test subject's ability are selectively given to each test subject,
An ability value calculating step for calculating the ability value of the examinee based on whether the answer of the item is correct;
An item selection step for selecting a predetermined item from an item database storing test items based on the calculated ability value;
A reliability coefficient calculating step of calculating a reliability coefficient of the capability value,
In the item database, items belonging to a plurality of item groups in which the order of questions is defined are stored,
The item selecting step includes a step of selecting an item belonging to the next item group according to the order of questions when the reliability coefficient of the ability value reaches a minimum value set for each item group. A test execution method characterized by that. When the item selection step asks a predetermined number of items or more in the certain item group, even if the reliability coefficient of the ability value does not reach the minimum value set for each item group, the question selection 8. The test execution method according to claim 7, further comprising a step of selecting an item belonging to the next item group according to the order. The item database stores a first item for inquiring about a first aspect and a second item for inquiring a second aspect,
The item selection step includes a step of selecting an item related to the second aspect when a reliability coefficient of the capability value related to the first aspect reaches a certain threshold value. 8. The test execution method according to 8. Even if the reliability coefficient of the capability value related to the first aspect does not reach a certain threshold value, the second item related to the second aspect is determined when the first time limit has elapsed. The method for performing a test according to claim 9, further comprising the step of selecting: In the item selection step, when the reliability coefficient of the capability value related to the second aspect reaches a certain threshold value, the reliability coefficient related to the first aspect is referred to, and if necessary, the first value The test execution method according to claim 10, further comprising the step of selecting: Furthermore, the display field generation step for generating an answer status display field for notifying the progress of the test on the display device of the terminal device for presenting the item and allowing the examinee to input an answer,
The test execution method according to claim 7, wherein the display column generation step includes a step of advancing a variable bar in the answer status display column as the item group progresses.

Images