JP4418693B2 - Group evaluation device - Google Patents

Description

The present invention relates to a group evaluation apparatus for evaluating a group composed of a team leader and one or more members .

  Companies are making various efforts to effectively utilize human resources in order to win the tough competition between companies. Effective use of human resources not only considers individual abilities and aptitudes, but it is also important to ensure that when a team of multiple individuals is formed, that team can perform at its best. When a project is carried out by a team, it is necessary to consider what kind of human resources should be gathered and what kind of human resources should be combined for the best team performance. Don't be. In the team organization method of simply selecting the ones with high abilities in order and putting them into the team, the compatibility between the leader and members, the compatibility between members, and the complementary relationship may not be taken into account, and it may not necessarily be the best team organization. is there. In addition, the elements that should be emphasized in team formation often differ depending on the goals and achievement levels required of the team.

In the previous technology, there was a technology for evaluating individual human resources (for example, patent application No. 2003-384391), but with this technology alone, what kind of human resources should be gathered to maximize the performance of the team It was difficult to find out what kind of human resources should be combined.

  This invention aims to solve the above-mentioned problems, and a group that derives what kind of human resources should be collected and what kind of human resources should be combined in order to maximize the performance of the team that is the group. The purpose is to provide evaluation technology.

In order to solve the above problems, the group evaluation apparatus according to the present invention makes it possible to evaluate a group (project team) composed of a team leader and one or more members.

The group evaluation apparatus according to the present invention evaluates a group composed of a team leader and one or a plurality of members, and the values of evaluation items measured in advance for each of a large number of team leaders. description stored as variable values, and for each even number of members, the independent variables value storage unit 101 stores the value of the evaluation items were measured in advance as independent variables value of each member, the plurality of team leaders and numerous a population structure table 102 from a member stores membership of a particular to each group of the population, for each group stored in said population configuration table 102, are stored respectively with their arrangement all members maximum value from the independent variables, minimum, mean, standard deviation, kurtosis, calculates the skewness, and team Li of the population From the independent variables value of loaders, generates data of the population independent variables values, the population described variable value storage unit 103 stores the, for a population stored in the population configuration table 102, for each population, success teams A group reference variable value storage means 104 storing values of reference variables previously evaluated as unsuccessful teams and intermediate teams , a group explanation variable value stored in the group explanation variable value storage means 103, and the group reference Based on the value of the group reference variable for each group stored in the variable value storage means 104, one or a plurality of first groups having a desired reference variable value by decision tree analysis, and a desired reference variable value A decision tree analyzing means 105 for classifying a group into a plurality of groups including one or a plurality of second groups that do not have, and obtaining a branching condition for making the classification, and to be evaluated Comprising a population independent variables value of the population, based on the classified branch condition in the decision tree analysis unit 105, a population of said evaluation subject, the first group having a value of the desired reference variable, or the It is characterized by having group evaluation means 108 for determining whether it belongs to any of the second groups not having a desired reference variable value and outputting the determination result.

  According to this aspect, it is possible to predict by a statistical method whether or not a certain group can achieve the expected result.

  According to this aspect, in order to realize the value of a certain reference variable, it is possible to statistically derive an optimal group configuration (which component is selected to configure the group).

According to the present invention, when a certain project is performed, it is possible to determine the suitability of a team composed of a leader and a plurality of members suitable for the project execution.
According to another aspect of the present invention, when a certain project is executed, it is possible to determine a combination of a leader and a plurality of members suitable for the project execution.

Embodiments of the present invention will be described below with reference to the drawings.
[Device configuration]
FIG. 1 is a functional block diagram showing a configuration example of an apparatus (hereinafter referred to as “group evaluation apparatus”) that evaluates a group of a plurality of constituent elements and evaluates a group according to a certain evaluation criterion. .

  The group evaluation apparatus 10 is an information processing apparatus such as a computer or a workstation, and includes an arithmetic processing unit (CPU), a main memory (RAM), a read-only memory (ROM), an input / output device (I / O), and As needed, it is composed of a device having an external storage device such as a hard disk device. A program for causing a computer to execute a program for causing the information processing apparatus to function as the collective evaluation apparatus 10 is stored in the ROM or an external storage device, and the program is loaded on the main memory, and the CPU stores the program. By executing this, the group evaluation apparatus 10 is realized. In addition, the above program does not necessarily have to be stored in a storage device in the information processing apparatus, and a computer-readable program recording such as a magnetic disk, an optical disk, a magneto-optical disk, a CD (Compact Disc), a DVD (Digital Video Disc), etc. The configuration may be such that it is provided from a medium or an external device (for example, an ASP (Application Service Provider) server), and this is loaded on the main memory.

  The group evaluation apparatus 10 includes a constituent element explanatory variable value storage unit 101 that stores a plurality of explanatory variable values for constituent elements that constitute a group, a group constituent table 102 that stores constituent elements of each group, and a constituent element explanatory variable. Based on the storage contents of the value storage means 101 and the population configuration table 102, a plurality of explanatory variable values for each group are generated and stored, and the group explanatory variable value storage means 103 for storing the values and the reference variable values for each group are stored. Based on the values stored in the group reference variable value storage means 104, the group explanation variable value storage means 103 and the group reference variable value storage means 104, one or a plurality of first variables having a desired reference variable value by decision tree analysis Group the measurement objects into a plurality of groups, including one group and one or more second groups that do not have the desired reference variable value. A decision tree analysis unit 105 that outputs the results and explanatory variables used for grouping; a branch condition extraction unit 106 that receives the output of the decision tree analysis unit 105 and extracts the grouping results and the explanatory variables used for grouping; A branch condition storage unit 107 for storing the result extracted by the branch condition extraction unit 106, a group evaluation unit 108 for evaluating the specified group based on the stored contents of the branch condition storage unit, and a specified reference variable And an optimum group determining means 109 for determining the constituent elements of the group so that it becomes the optimum group.

  The component (also referred to as an individual or a sample) handled by the group evaluation apparatus 10 may be any object that can be evaluated, and is not limited to a person. The constituent elements may include different types of elements. For example, the team leader and the constituent members can be constituent elements. The “group” is composed of a plurality of components, such as a company department, a project team, a sports starting member, a school class, and the like.

  Each component must have a value for the explanatory variable. “Explanatory variable” is an explanatory variable for classifying components. For example, if the component is a member that constitutes a group, it is difficult to grasp the original values such as “sociality” and “aggressiveness”. Any element that can be displayed numerically, such as a numerical element, “numerical processing capability”, and “sales” can be used as an explanatory variable.

  On the other hand, the value of the standard variable needs to be measured for “population”. “Standard variable” is information indicating the result of evaluating a measurement object against a certain evaluation criterion, for example, a successful team, a non-successful team, or a successful team, a non-successful team, For example, an evaluation of an intermediate team (a team that is not classified as a successful team or an unsuccessful team), and an evaluation of high, low, or high, medium, or low sales of the sales team.

  2 and 3 show examples of data stored in the component explanation variable value storage means 101. FIG. In this example, the group is a project team, the constituent elements are project team members and leaders, and the constituent element explanatory variable value storage means 101 stores both the explanatory variable values of the members and the explanatory variable values of the leaders. . FIG. 2 is an example of explanatory variable values for members, and FIG. 3 is an example of explanatory variable values for leaders.

  As shown in FIG. 2, the explanatory variable value data 200 for each member has one record 200R for each member, and each record is an evaluation variable “communication ability” 201, “variety of viewpoints” which is an explanatory variable. "202", "Innovation" 203, "Desirable / thrilling" 204, ..., the values are stored. In this example, the values of evaluation items measured by questionnaires and tests for each member are converted into deviation values and stored.

  Similarly, as shown in FIG. 3, the explanatory variable value data 300 for the leader has one record 300R for each leader, and each record is an evaluation item “Explanation variable” that represents the characteristic of the leader. The values are stored for “power to go” 301, “charisma” 302, “sympathy” 303, “fairness” 304,. As in the case of members, the values of evaluation items measured by questionnaires and tests for each leader are converted into deviation values and stored.

  The group composition table 102 stores the components of each group. For example, when the group is a project team, the group composition table 102 stores the leader and member (s) of each project team.

  An example of data generated and stored by the group explanation variable value storage means 103 is shown in FIG. The data 400 generated and stored by the group explanatory variable value storage means 103 has one record 400R for each project group which is one group. Each record stores a member number 401, a member value 402, and a leader value 403. The member number 401 is the number of members constituting the group. The member values 402 are “maximum value”, “minimum value”, “average”, “standard deviation”, “kurtosis”, and “distortion” in the values of all the explanatory variable members for each explanatory variable (evaluation item) of the constituent element. Are calculated and stored. The leader value 403 stores the explanatory variable value of the leader of the group.

  The group reference variable value storage means 104 stores the value of the reference variable of each group. For example, if the project group that is the group is a design group, the quality of the design drawing created by the group (high / medium / low), process control Items such as accuracy (high, medium, low) are standard variables. The value of the reference variable is generally a binary value such as success / non-success, and a ternary value such as high / medium / low.

  The decision tree analysis means 105 is an analysis for finding conditions for classifying a group into success / non-success or high / medium / low from the stored contents of the group explanatory variable value storage means 103 and the group reference variable value storage means 104. I do. Specific contents will be described later.

[Device operation]
Next, an operation example of the group evaluation apparatus 10 will be described. The group evaluation apparatus 10 operates to execute the branch condition acquisition method shown in FIG.

  First, in order to execute the branching condition acquisition method, the group evaluation apparatus 10 acquires a plurality of explanatory variable values for the leaders and members that are the constituent elements (step S601) and stores them in the constituent element explanatory variable value storage means 101. Let This is performed, for example, as an input to the computer of the result data of the aptitude test of leaders and members as constituent elements.

  Next, for all project groups that are groups, the leaders and members are registered in the group composition table 102. From the data stored in the group composition table 102 and the component element explanation variable value storage means 101, FIG. Data of group explanation variable values as shown is generated (step S 602) and stored in the group explanation variable value storage means 103.

  In addition, the group evaluation apparatus 10 acquires the value of the reference variable for the project group that is each group (step S603), and stores it in the group reference variable value storage unit 104. This is performed, for example, as an input to a computer of evaluation data scored by an evaluator of a project group, for example, an officer in charge.

  The group evaluation apparatus 10 includes one or a plurality of values having a desired reference variable value by decision tree analysis based on the data stored in the group explanatory variable value storage unit 103 and the data stored in the group reference variable value storage unit 104. The group is classified into a plurality of groups including the first group and one or a plurality of second groups that do not have a desired reference variable value (step S604).

  Next, the group evaluation apparatus 10 acquires a branch condition from the decision tree generated in step S604 (step S605).

  As described above, the group success / non-success or success / unsuccess / intermediate distribution information corresponding to the value of the group reference variable, and individual group explanation variable values that are values of the group explanation variable (FIG. 4). The branch condition is acquired using a statistical mathematical analysis technique called decision tree analysis.

  There are various methods such as CART, C4.5, C5.0, and CHAID for decision tree analysis. In this embodiment, for example, a method called CHAID (chain) analysis included in AnswerTree, which is a program product of SPSS, is used. Use. However, the present invention is established even if a decision tree analysis method other than the CHAID analysis is used, and the present invention is not limited to the CHAID analysis.

  The internal algorithm of CHAID analysis uses a widely used statistical mathematics method called chi-square test, and the success / unsuccessful or successful / unsuccessful / intermediate distribution information (= population reference variate value) and the collective explanatory variable The relationship between values is analyzed in detail.

  That is, for a decision tree analysis program such as CHAID analysis, the success / non-success of the group, or the distribution information of success / non-success / intermediate information on the reference variable and the explanation of the group generated from the explanatory variable values of the constituent elements of the group By inputting the variable value, the success / failure branch condition of the group is output.

  At this time, the value of the group explanatory variable is normally given as a deviation value (if it is not a deviation value, it can be converted into a deviation value), but when acquiring a branch condition, it is converted into three types of values as follows. That is, a value greater than the deviation value 54.31 is classified as “strong”, a value between the deviation values 45.69 and 54.31 is “medium”, and a value less than the deviation value 45.69 is classified as “weak”. Here, the deviation values “45.69” and “54.31” are deviation values corresponding to three equal divisions when the sample follows a normal distribution. Further, the condition values for these divisions may be changed according to options.

  A success model output from a decision tree analysis program such as CHAID analysis is a condition for success specific to the reference variable, and indicates a combination of group explanatory variables. The combination is called a success pattern, and is usually output in one or more ways. A set of success patterns is called a success model.

  The diagram shown in FIG. 5 is a diagram showing an example of the output of CHAID analysis corresponding to the decision tree analysis step (output of the decision tree analysis means 105). Nodes (N0 to N11) indicating a set of project groups and group explanation variables (C1 to C5) adopted when classifying the upper layer nodes into a plurality of nodes are displayed. Further, in the box indicating each node N0 to N11, the reference variable (evaluation High / evaluation Low or evaluation High / evaluation Low corresponding to success / non-success, or success / unsuccess / intermediate) of the employees belonging to that node is displayed. / Evaluation Middle distribution information). In this example, when the ratio of the group given the evaluation High included in the node is equal to or greater than a predetermined value, the node is determined to be a successful pattern.

  In the example shown in this figure, of the success patterns A (node N6), B (node N11), C (node N9), and D (node N4) among the nodes N1 to 11 branched from the node N0 which is the population. Four success patterns are output. For example, the success pattern A branches in order from the box labeled “N0” at the top to the node 2 (N2) and the node 6 (N6) in order, and the combination of the branch conditions is “the power to face the difficulties of the leader” The deviation value of “C1” is “medium or strong”, and the deviation value of “maximum viewpoint diversity” (C3) is “strong”. Based on this branch condition, the branch condition of the success pattern A corresponding to one of the first groups having a desired reference variable value is acquired.

  The other successful patterns B, C, and D are also output with combinations of branch conditions.

  The information written in each box called a node indicates how many project groups are allocated to Evaluation High / Evaluation Middle / Evaluation Low and what percentage of the entire node. In the nodes (N6, N11, N9, and N4) corresponding to the success pattern, the evaluation high group is larger than the evaluation middle / evaluation low group.

Returning to FIG. 1, the description of the configuration example of the apparatus 10 is resumed.
From the analysis result of the decision tree analyzing unit 105, the branch condition extracting unit 106 extracts the branch condition and classification (success pattern, unsuccess pattern, or success pattern, intermediate pattern, unsuccess pattern) for each node in association with each other, It is passed to the branch condition storage means 107. The branch condition storage means 107 stores the branch condition and classification for each node. The stored branch condition and classification for each node are used by the group evaluation means 108 or the optimum group determination means 109 described later.

[Operation example of group evaluation tool]
Next, an operation example of the group evaluation unit 108 will be described with reference to FIG.
First, the operator of this apparatus specifies the constituent elements of the group to be evaluated using input means such as a keyboard (step S701). For example, the operator inputs the ID of the leader of the group and the IDs of all members of the group and passes them to the group evaluation means 108.

  The group evaluation unit 108 acquires the value of the explanatory variable for each component (step S702). If the value of the explanatory variable of the constituent element has already been stored in the constituent element explanatory variable value storage means 101, the group evaluation means 108 reads and acquires it. If not stored in the component explanatory variable value storage means 101, the operator is requested to input the explanatory variable value of the component.

  Next, the group evaluation unit 108 generates and stores a group explanatory variable value for the group (step S703). This generation method is the same as that performed by the group explanation variable value storage means 103.

  Next, the group evaluation unit 108 compares the stored group explanatory variable value with the stored contents of the branch condition storage unit 107, and the group to be evaluated belongs to which node of the tree as shown in FIG. Is determined (step S704).

  Finally, the group evaluation unit 108 outputs a determination result (step S705). For example, the group evaluation unit 108 outputs a comment that “this group X belongs to the node N6 that is the success pattern A” or a diagram in which a mark is added to a node that is determined to belong to the group.

  With this output information, the user can statistically predict whether or not the group to be evaluated is a successful group.

[Operation example of optimal group determination means]
Next, an operation example of the optimum group determination unit 109 will be described. First, the optimum group determination unit 109 creates all possible group configurations. For example, there are _N C _r combinations that create a group of r people from N component candidates, and the optimum group determination means 109 creates a group of all _N C _r component combinations.

  Next, the optimal group determination unit 109 performs the processing of steps S701 to S705 in FIG. 7 for all the generated groups in the same manner as the group evaluation unit 108, and acquires the determination result of each group.

  Finally, the determination result of each group is compared, and the group that has obtained the highest determination result is output as the optimum group. As a result, the user can statistically obtain a group having an optimal configuration (organization) for a certain project.

Functional block diagram showing a configuration example of a group evaluation device The figure which shows the example of the memory content of a component description variable value storage means The figure which shows the example of the memory content of a component description variable value storage means The figure which shows the example of the memory content of a group explanatory variable value storage means Figure showing example output of decision tree analysis means Flow chart to branch condition output Flow diagram for outputting group evaluation results

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Group evaluation apparatus 101 ... Component element explanation variable value storage means 102 ... Group structure table 103 ... Group explanation variable value storage means 104 ... Group reference variable value storage means 105 ... Decision tree analysis means 106 ... Branch condition extraction means 107 ... Branch Condition storage means 108 ... group evaluation means 109 ... optimum group determination means N0 to N11 ... nodes C1 to C5 ... branch conditions (explanatory variables)

Claims (1)

A group evaluation device for evaluating a group composed of a team leader and one or more members,
This group evaluation device
For each of a large number of team leaders, pre-measured evaluation item values are stored as explanatory variable values for each team leader , and for each of a large number of members, pre-measured evaluation item values are also described for each member. Explanatory variable value storage means 101 stored as a variable value;
A group composition table 102 that identifies the group from the plurality of team leaders and members and stores the members of each group;
For each group stored in said population configuration table 102, and calculates the maximum value from the independent variables value stored respectively with their arrangement all members, minimum, mean, standard deviation, kurtosis, the skewness, And, from the explanation variable value of the team leader of the group , data of the group explanation variable value is generated, and the group explanation variable value storage means 103 storing this data ,
For the groups stored in the group composition table 102, for each group, a group reference variable value storage means 104 that stores reference variable values that have been evaluated in advance as a successful team, an unsuccessful team, and an intermediate team ;
Based on the group explanation variable value stored in the group explanation variable value storage means 103 and the value of the group reference variable for each group stored in the group reference variable value storage means 104, a desired standard is determined by decision tree analysis. Classifying a group into a plurality of groups, including one or more first groups having a variable value and one or more second groups not having a desired reference variable value, and forming the classification Decision tree analysis means 105 for obtaining a branch condition;
Have
Based on the group explanation variable value of the group to be evaluated and the branch condition classified by the decision tree analysis means 105, the group to be evaluated has the first group value having the desired reference variable value. Or a group evaluation unit 108 for determining whether it belongs to any of the second groups not having the value of the desired reference variable, and outputting the determination result.

Images