JPH0997250A - Questionaire processing system and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a means saving the manpower in the time from the preparation of a questionaire to the preparation of a data analysis and improve the efficiency by the reuse of questionaire know-how and questionaire data, regarding a system and method therefor preparing questionaire and processing the data of result of the questionaire carried out based on this questionaire. SOLUTION: This system is provided with a question structure information input means 1 inputting the question control sentence describing the structure of a question according to the question type based on a specified notation method and storing the sentence in a question structure information storage part 2 by a dividing the sentence for every question, a questionaire pattern storage part 4 storing the questionaire display form corresponding to the question type and a questionaire preparation means 3 preparing the questionaire for every question based on the question structure information stored in a question structure information storage part 2 and the questionaire display form of the questionaire pattern storage part 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for creating a questionnaire for a questionnaire survey and processing data of the results of a survey carried out based on this questionnaire.

[0002]

2. Description of the Related Art Conventionally, work such as creating a survey form, creating a punch format for inputting a survey result, data logic check, and data analysis have been performed individually for each stage. In addition, there were many parts where each work was done manually by trial and error.

[0003]

The problems at each work stage are as follows. 1. Labor saving in clearing work when creating survey forms Until now, when preparing survey forms, each question was written one by one with a word processor, etc., and it was created by trial and error while observing the layout on the screen. In addition, when you want to add a question or change the order of questions after it is completed, you need to change the question number and reconfirm all the contents.

Design of Survey Form for Prompting Correct Answer If the user is inexperienced in designing a survey form, it was difficult to design an easy-to-understand survey form that prompts the respondent to answer correctly.

Design and instruction of data input specifications After collecting the survey form, there is a work of punching the data, but the survey form creator also had to create the punch format. 2. At the time of analysis Checking the logical consistency of answers Currently, SAS, SPSS, ANAL are used for data analysis.
Commercially available analytical tools such as YST are used. When the number of questions is large, it is difficult to create these input sentences and the system needs to generate them automatically.

[0006] In addition, since the person in charge appropriately assigns the variable name and label to each question, it is difficult to reuse the data. Information such as who analyzed what question and what structure the question was used for should be stored in a form that allows quick understanding.

In view of the above points, the present invention provides a means for realizing labor saving of work from the preparation of a survey sheet to the start of data analysis, survey know-how, and efficiency improvement by reusing survey data. To aim.

[0008]

The above problems can be solved by a survey form processing system configured as follows.

FIG. 1 is a block diagram of the present invention. Reference numeral 1 denotes a question structure information input means for inputting question structure information in which a question structure is described based on a notation defined for each question type and storing the question structure information in a question structure information storage unit 2 for each question.
Is a survey form pattern storage unit that stores a survey form display format corresponding to the above question type.

Reference numeral 3 is a survey form creating means for creating a survey form for each question based on the question structure information stored in the question structure information storage unit 2 and the survey form display format in the survey form pattern storage unit 4 ( Claims 1 and 4).

Reference numeral 5 denotes a data structure information analysis means for extracting information on the structure of the answer data to the question from the question structure information for each question stored in the question structure information storage unit 2 and storing it in the data structure information storage unit 6. Yes, 7 is based on the data structure stored in the above data structure information storage unit 6,
This is a survey data input means for inputting reply data for the survey sheet created and storing it in the survey data storage section 8 (claims 2 and 5).

Reference numeral 10 is a totalization processing pattern storage unit that stores the totalization processing format corresponding to the above question type, and 9 inputs the survey data stored in the survey data storage unit 8,
It is a totalization processing means for totaling for each question based on the totalization processing format stored in the totalization processing pattern storage unit 10 corresponding to the question type of the question structure information storage unit 2 (claims 3 and 6).

[0013]

2, 3 and 4 are explanatory views (No. 1) to (No. 3) of the configuration of the embodiment of the present invention.

First, the structure will be described. The same reference numerals are given to those corresponding to the configurations of the present invention. FIG.
An explanatory view (No. 1) of the configuration of the embodiment of the present invention corresponds to claim 1 and claim 4, and is used for inputting a question control sentence to obtain question structure information and creating a questionnaire. The configuration is shown.

Reference numeral 21 is a device for inputting question control sentences,
From the display screen, etc., the specification of the question of the questionnaire to be created is input as a question control statement. Reference numeral 1 denotes a question structure information input unit, which stores the question control text input from 21 into the question structure information storage unit 2 as the question structure information in a format determined for each question, organized by corresponding items. ing. The format defined above is 22
It is stored as a table in the question structure format table storage unit.

The question structure information storage unit 2 stores all the elements that determine the structure of each question.
The contents of processing at each stage of the survey result data based on the survey form after the creation of the survey form will be determined.

At the stage of creating the questionnaire, the questionnaire creating unit 3 extracts the question structure information for each question stored in the question structure information storage unit 2 and prints or displays it stored in the questionnaire pattern storage unit 4. The pattern corresponding to the question type recorded in the extracted question structure information is selected from among the patterns of the questionnaire form indicating the format, and the questionnaire 23 is created based on the content of the question structure information.

The explanatory view (No. 2) of the configuration of the embodiment of the present invention in FIG. 3 corresponds to claims 2 and 5, and is extracted as data structure information based on the question structure information, The structure for preparing the survey data so that various survey results can be collected by inputting the survey results collected by conducting the survey based on the survey sheet based on the above data structure information is shown.

The data structure information analysis unit 5 extracts the question structure information from the question structure information storage unit 2, analyzes the data structure for each question, and stores it in the data structure information storage unit 6 for all the questions.

This data structure information is the data structure list 3
It will be described later by referring to the data structure list 31 as well as using it as the format for inputting the information entered in the questionnaire collected as the survey result and output as 1 as punch data. It can be used to reuse and analyze the data of the survey results.

The survey form 32 collected by conducting the survey corresponds to the data structure stored in the data structure information storage unit 6, and the survey data input unit 7 has the format shown in the data structure list 31. In order to handle the data of the questionnaire with punched data of the questionnaire with various analysis tools, the data can be divided and organized based on the contents of the data structure information storage section 6, and the names can be associated with the data. And stores it in the survey data storage unit 8.

FIG. 4 is an explanatory view (No. 3) of the configuration of the embodiment of the present invention, which corresponds to Claims 3 and 6, and is for carrying out the tabulation processing of the survey data of the collected survey sheets. The configuration is shown.

Based on the question structure information of the question structure information storage unit 2, the data structure information of the data structure information storage unit 6 and the survey data of the survey data storage unit 8, the form of the aggregation process is stored corresponding to the question type. Totalization processing pattern storage unit 10
The aggregate processing pattern corresponding to the question type indicated in the question structure information is extracted from the data structure information storage unit 6 to know the structure and location of the survey data for each question, and the corresponding survey data in the survey data storage unit 8 is input. The totalization table 41 in the format indicated by the totalization processing pattern is output.

Next, each part of the embodiment of the present invention will be described in detail. FIG. 5 shows an example of the question control sentence.
The question control sentence that describes the question of the questionnaire of the questionnaire survey is designed to show the content of each question by a predetermined combination of symbols.

The four questions (shown as question and Q) are shown separately in (a), (b), (c), and (d).
Subquestion (subquestion, SQ) is shown in (a).
Includes two. Further, although there are questions not shown here between (b) and (c), the following description will be given assuming that (a) to (d) are arranged in the order shown here and input as question control statements.

In (a), Q is "What is your gender?"
Asked the gender of the respondent. \ Indicates that the alphabetic character following in the control statement has a special meaning, and \ Q and \ SQ mean that a question or a subordinate question starts. \ T represents a question type and is indicated by the alphabet shown subsequently. SA here means to select only one from the options arranged in the meaning of single answer, that is, to mark the corresponding number. (Sex) and "sex" are necessary when referring to or displaying the data of the survey results, and indicate the variable name and label.

"\ MQ 2" indicates that two subordinate questions are attached to this question. As a result, two SQs are arranged next to each other. \ C (1) indicates an option that is an answer in the meaning of the category. That is, "male" and "female" are shown. (1) after \ C indicates what number of the next SQ should be answered when the option is answered. If you answer "Men" here, SQ1
Answering "female" means instructing to proceed to SQ2.

In the next SQ, as a question for men, it is required to answer one of the choices "car", "bicycle", and "bike" as the "most wanted item". The next SQ specifies that women can choose as many "interests" as they want from "overseas travel", "movies", and "gourmet". \ TMA [*] indicates the question type of multi-answer, and when there is a number in [], it indicates the number of options to be answered, but as in this example, it is "*" instead of a number. In this case, show that you can select and answer any of the options shown in the question.

FIG. 6 shows an example of the survey form. The questionnaire corresponding to question 1 above is shown in (a). There are the question text of question 1 starting with Q1 and the options enclosed in a square frame, as well as the subordinate question texts starting with SQ1 and SQ2.
Under the square frame of Q1, "If you answered male, go to SQ1,
If you replied that you are a woman → Go to SQ2 ", but \ C
It was created corresponding to (1) and \ C (2), and "only one" is added to the question text of SQ1 and SQ2.
The "any number" part is underlined. These are automatically generated to get proper answers.

Similarly, in FIG. 5B, \ T MA [3]
As shown below, there are 6 options from Company A to Company 6
It indicates that only three companies are selected from the company names of the companies. The form of the questionnaire is shown in Q2 of FIG.

In the right side of each question, as shown in the example of the questionnaire form of FIG. 6, is displayed as punch data using the data of the survey result, that is, the number of the choices of the questionnaire that answered each question as punch data. Indicates the digit position to perform. This is because the number of reply data changes depending on the question structure for each question and is recorded as a data structure and automatically output to the survey form.

5 (c) and 5 (d) show examples of question types (successive categories, numerical entry) different from those described above, which will be described later. FIG. 7 shows question structure information elements classified by question type. This indicates what kind of symbol is specified for each question type.

The type of question type is SA (single answer), MA (multi answer), V (numerical entry), MT (matrix), SD as indicated in the question type symbol column.
(Successive category), J (rank data), etc.

The symbols to be designated include the number of selections, the number of steps, the unit, the variable name, and the label as shown in the option symbol column, and the symbols are [], {}, <>, (), respectively.
It is specified by letters or numbers enclosed in parentheses such as "".

Category designation symbols are C, RO, CO
Shows the answer character string which is an option. In particular, RO indicates options shown in horizontal rows in the matrix format, and CO shows options shown in vertical columns.

I is a small question symbol in matrix form,
It is an element of the question that is configured in the case of successive categories and ranking data. "○" and "-" shown in each column of the table
Indicates whether or not each symbol is applied to each question type.

For example, "variable name" and "label" are symbols that can be specified in any question type, but the unit shown in <> is specified only when the question type is V (numerical entry). Show. There is a question to answer the numerical value of the age in (d) of Fig. 5, but "\ T V [2.0] (ag
e) Used to indicate the age unit “age” as in “age” <age> ”.

Next, FIG. 8 shows a question structure format table.
The question structure format table is stored in the question structure format table storage unit 2.
The question structure information storage unit 2 is a template stored in advance in FIG. 2, and the question structure information is organized and stored for each question according to the format of this table.

The structure of the question structure format table will be described with reference to FIG. Roughly speaking, the left and right columns are divided into question structure information element names and information formats. The question structure information element name indicates the element of the question structure for each question, and is a question number indicating the position in the question group of the question in addition to the various symbols specified in the question control statement in FIG. 7 described above. "Question counter 1", "Question counter 2" that shows the number of the sub-question (subquestion), "Start column position" that shows the digit position where the answer of this question is punched, Question sentence itself, "Subquestion branch destination designation" There is.

There are three types of information formats in the right column. "Numerical type" is a short square frame indicating a numerical value such as the number of selections, "short text" is a medium square frame indicating a character string such as an inquiry symbol, and "long". "Text" is indicated by a long rectangular frame indicating a long character string such as a question sentence. ... on the right of the long rectangular frame indicates that there can be multiple designations.

The small square box at the lower right is used to numerically indicate the number of designated elements for a plurality of elements that can be designated. Examples of question structure information are shown in FIGS. 9 and 10.

FIG. 9 shows an example of the above question, which corresponds to the question of FIG. 5 (a). The question counter 1 that is the first question and indicates the question number is "1". The start column position is the first data and is "1" as shown in the first column.
It is.

The question symbol is an ordinary question and is Q. The question text is "What is your gender?" Question type is "SA". As for the option symbol variable name, "sex" is specified in (sex) as in "\ TSA (sex)" sex "\ MQ2" of the question control statement. The option symbol label is “gender” and “gender” is written as described above.

The category designation symbols are “male” and “female”
Is specified as an option, and both are recorded with the specified number "2". The number of sub-question branch destinations is recorded as "2" like "\ MQ 2". The branch destination is a long rectangular frame, which means SQ1 for male and SQ2 for female, as designated by "\ C (1) male \ C (2) female" in FIG. "1" and "2" are recorded.

FIG. 10 shows the question structure information corresponding to the example (b) of the question control statement of FIG. 5 in the example (part 2) of the question structure information. The question number is SQ in the sequence of question control statements as shown in Fig. 5.
Since the question counter is counted excluding, the question counter becomes "2".

The starting column is 1 digit for Q1, 1 digit for SQ1,
Since SQ2 uses 5 digits in total of 3 digits, it indicates that Q2 consists of digit number 6. Question type is "MA"
Indicates a multi answer. The number of options to be answered is specified as [3], which is "3".

Six category designating symbols, \ C company to \ C company, are designated, which are also written in order in a long rectangular frame, and the designated number is "6". As described above, the symbols designated in the question control statement are sorted and stored according to the template in the question structure format table storage unit 22 for each question and stored as question structure information.

This information is commonly used at each stage of data structure information analysis for determining data input specifications from survey sheet preparation and survey data analysis. Next, description will be given of the survey form creating unit 3 that creates the survey form 23 using the question structure information stored in the question structure information storage unit 2 and the survey form pattern stored in the survey form pattern storage unit 4. to go into. First, the survey form pattern will be shown.

FIG. 11 shows the structure of the survey form (part 1).
Is shown. Here, question types SA, MA, V, and J are shown. For SA (single answer), the questionnaire pattern is displayed as Q □ or SQ □, followed by the text of the question text. Column numbers are shown on the right shoulder. The options specified in the category are 1, 2, 3 at the beginning.
-Numbers are attached in order to indicate the answer items of the options specified by \ C, and they are enclosed in a square frame.

When MQ (multi-question) is designated, it indicates the SQ number to be advanced next in correspondence with the option. In the case of MA (multi-answer), it is almost the same as SA, but since multiple answers are requested, multiple column positions of punch data are designated and the format is □ to □. In addition, SQ
Is only for SA questions.

In the case of V (numerical entry), the format shows the arrangement of numbers and the position of the decimal point instead of options. Present a square box. In the case of J (sequential data), a square frame in which a number can be entered from the options is presented in order from the first place.

FIG. 12 shows the structure of the survey form pattern (No. 2)
It is. The question types of SD and MT are shown below. SD
In the case of (successive category), only the vertically designated stages of the "items indicating the degree" that are options are shown, and correspondingly, the numbers indicating the stages are 1, 2, 3, ... Line up. One number indicating the stage is answered for each small question, so a column position column is also prepared for each question. See FIG. 5C for a specific example of the SD type question control statement.

In the case of MT (matrix), it is a format for answering the question item for each column indicated by CO for each item of the row indicated by RO, and the options for the row and column matrix are 1, 2, 3 as shown in the figure. 4 is presented and selected.

FIG. 13 shows the structure of the emphasized sentence pattern. The question text of the questionnaire should be written in an easy-to-understand manner in order to obtain the intended answer, but in order to get the necessary and sufficient number of options to answer, The wording indicating the number of responses is automatically emphasized and marked to improve the quality of the response data of the questionnaire.

Therefore, in the embodiment of the present invention,
In particular, a keyword that requires special attention is stored in advance in the survey form pattern storage unit 4 together with the survey form pattern, and the keyword is collated with the question text. Specifically, an operation of underlining the character string portion that matches the keyword is performed.

The operation of creating the survey form in the above-mentioned configuration will be described. FIG. 14 is a flowchart of the survey form creating unit. In step S141, the question structure information is sequentially extracted from the question structure information storage unit 2. The question structure information is stored for each question, and one question is created for each question.

In step S142, the question sentence is emphasized. Information on the question sentence is extracted from the question structure information, and character string comparison is performed for each of the keywords of the emphasized sentence pattern in the survey form pattern storage unit 4 described above. If there is a match, the underline information is added to the character string in the question text.

Next, in step S143, the survey form pattern corresponding to the question type of the question structure information is retrieved from the survey form pattern storage unit 4. In step S144, one question of the questionnaire is formed by embedding the constituent elements of the question structure information corresponding to each part in the retrieved questionnaire pattern in the display printing format of the questionnaire. That is, since the question number is sequentially counted in the question counter 1 in the order of the question sentences, it is used as the question number Q ... If necessary, use the question sentence that has been emphasized. As the column position, the column position of the data punch, which is also set as the start column position in the question structure information, is embedded in the position to be displayed and printed. Move the answer character string that is the option specified in the category to the answer item position. In addition, depending on the survey form pattern, the answer item part may be surrounded by a square frame.

Formally, the display print format is a vertical character string, and the number of items changes depending on the question, but the arrangement and embedding of information corresponding to this is not described here because it corresponds to the conventional technique.

In steps S145 and S146, subquestion (SQ) is performed only when the question type is SA.
Since there is a possibility that there is such a designation, the item of the number of sub-question branches in the question structure information is checked.

When the question is followed by subquestions, a branch destination directive is constructed in step S147, and a subquestion number to be advanced is guided between the subquestion and the next subquestion (SQ) depending on how to select an option. The number of sub-question branches is in the question structure information. Examine as many sub-question branch destination specified long texts as there are, and specify the branch destination SQ number specified in the question structure information category. It is constructed as a sentence that indicates the SQ number to be advanced corresponding to the symbol, that is, the answer item character string of the option. As a specific example, as shown between Q1 and SQ1 in Fig. 6, those who answered male → SQ1 answered female → SQ2 “Male” stored in the category designation symbol
It is created by associating "female" with "1" and "2" stored in the sub-question branch destination designation and embedding them in a standard sentence.

In step S148, it is confirmed whether or not the processing has been completed for all the question structure information. If there is any remaining, the process returns to the beginning and the processing is repeated for the next question structure information. Next, the operations of the data structure analysis unit and the survey data input unit shown in FIG. 3 will be described.

In order to perform the statistical analysis processing of the survey form collection data that answers the questions, it is necessary to store it as a data file in a form that can be handled by various analysis tools. The survey form may be directly captured as electronic information from a terminal device, may be collected through an electronic medium, or may be read by an input device such as OCR or OMR. In the description of the embodiments of the present invention, explanations are made assuming that a questionnaire is displayed on a paper or an image of a questionnaire image is displayed, but the statistical analysis of the results can be performed no matter what medium the answers to the questions are collected. The format of the analysis data required when performing is determined by the question structure and does not depend on the type of medium. Here, we will use the ones marked on the most common survey form as the source data.

First, FIG. 15 shows an example of the data structure information. The information of the data structure is all the information related to the structure of the survey data extracted and arranged from the question structure information in the question structure information storage unit 2 described above. Some names of data are regularly created from these.

One of the data structures is information required as a relation with the input medium, and in the input image to the floppy medium or the like, reference numeral 151 in FIG.
There is a "column" shown in. Set the first digit position to "1",
It is illustrated that there is input information up to “80”. These values are transcribed from the starting column position of each question structure information.

Reference numeral 152 is a question number, Q1, Q2, ...
.. The subquestion indicates that it is subordinate to the question, and Q1S1 and Q1S2 indicate SQ1 and SQ2 following Q1. Reference numeral 153 is a question type, which depends on the question type of the question structure information.

Numeral 154 denotes the number of selections, which is the number of columns determined for each question type. As shown in the flowchart described later, for example, when the type is MA, it is the number specified by the option symbol [] of the question structure information, and when the type is V, it is the integer part of the number in []. It is like this.

Reference numeral 155 is a variable name. If there is a variable name indicated by an option symbol () in the question structure information, it is as it is,
If not, it will be created from the question number. For example, q1s1, if there are multiple data in the question, add a serial number.

Reference numeral 156 is a label, which is designated by the option symbol "" of the question structure information. 157 is a file name of the survey data when a set of data extracted for each question as the survey data is stored as a file. This name is also created from the question number.

The beginning part 158 of the question is obtained by cutting out the beginning part of the question sentence and recording it as a reference. In this way, the data structure information is stored in the data structure information storage unit 6.

FIG. 16 is a flow chart of the data structure information analysis unit 5 which creates the above data structure information by extracting and analyzing it from the information in the question structure information storage unit 2. The operation will be supplementarily described with reference to this flowchart.

In step S161, the question structure information is sequentially extracted for each question from the question structure information storage unit. Step S
In 162, the column number is obtained from the information of the starting column position.

In step S163, the calculation of the end column number in the question which answers a plurality of data is shown. As described above, the number of reply data is determined by the symbol designated as the question type for each question, and the number of columns is also determined.

As an example, MT1 (matrix type)
In the case of, the number of ROs specified is equal to the number of COs specified, but there is only the number of reply data obtained by multiplying the category symbol (RO) in the question structure information by the number designated in the category symbol (CO). Shows.

As described above, in step S164, the number of selections in step S165 is the above-described column addition number except V, and in the case of V, the numerical values are answered in a sequence of numbers and the number of selections is meaningless.

Steps S166, S167, S168, and S169 have been described above. One record of the data structure information created for one question in step S1610 is output to the data structure information storage unit 6.

In step S1611, it is confirmed whether all the question structure information has been processed. If not, the process returns to the beginning and continues processing for the next question. The data structure information created in this way is first printed and used as the data structure list 31. One of the uses is to clarify the column position instruction of data for each question as an input specification when punching data as described above. Also, when analyzing the collected survey data as described above, it is possible to know what sort of survey contents the survey data includes, and how variable names and labels are attached. Sharing and reuse can be done efficiently.

FIG. 17 shows an example of the format of survey data.
In (a), only the main points are shown in order to show an example of the response on the questionnaire form. (B) shows the correspondence with the column position of punch data.

(C) is a diagram in which the above data is divided into files as survey data for each question. (A) is illustrated as an example of a response sheet for the survey form partially shown in FIG. 6 created by the question control statement of FIG. The numbers in circles are the response marks. Further, the square frame in Q25 corresponds to the question control sentence in FIG. 5 (d) and is the answer to the question asking the age, and the answer is 1 and 9 is 19 years old.

(B) corresponds to (a), and what is indicated by a square frame is punch data. Correspondence with the column position number is shown above the square frame and the question number in (a) is shown below.

In (c), the data extracted for each question are arranged, and the one shown above is the file name of the survey data. Also, what is shown in the right is a group of data for questions that answer multiple data.

FIG. 18 shows a flowchart for inputting survey data. The data structure information in the data structure information storage unit 6 created above is sequentially taken out and is collected and output as a file as survey data for statistical analysis for each question.

In step S181, the data structure information for one question is taken out, and in step S182, the file for outputting the survey data for one question is opened. At this time, the output file name is the one specified in the extracted data structure information (survey data).

In step S183, one record of the survey form punch data is read. This is as shown in FIG. In step S184, punch data corresponding to the column position in the extracted data structure information is extracted, and in step S185, it is checked whether the question type is “V”, which is handled differently.

At step S186, since the question type is V, the data of a plurality of columns of punch data is converted into numerical information and output as one numerical data. FIG.
The data at the beginning of q25data in (c) is (a),
The sequence of numbers corresponding to Q25 in (b) is the numerical data "1.
You can see that it is 9 ”.

In step S188, if the question type is other than V, it is converted into numerical information column by column and sequentially output as the selected number of data. In q2data in FIG. 17C, 2, 5 and 6 are respectively output as one data as indicated by the circles in 2, 5 and 6 in (a).

In step S187, it is confirmed whether all the data structure information has been processed, and in step S189, the investigation data file is closed. Next, the explanation of the simple totalization process will be made. FIG. 19 shows an example (1) of a simple answering process for a single answer question, and FIG. 20 (2) shows a case of a multi-answer question.

FIG. 19 shows a pattern of a simple tabulation process corresponding to a single answer (SA) in which the question of FIG. This aggregation processing pattern is shown in FIG.
Are stored for each question type in the tabulation processing pattern storage unit 10 of FIG. "gender male Female",
The square frame in which "7", "5", and "7" are written is a part that changes depending on the question and the survey result. Given these changing information, this simple tabulation pattern calculates the cumulative frequency,% and cumulative% and displays the numerical values as a table. Further, the bar graph of the frequency is represented by scaling so as to include the maximum value of the given frequency. Here, there are two choices, "male" and "female", but as shown in the left side grouping, the number of frequency data and the number of graphs are changed depending on the number of choices, so single Simple aggregation about answers can be covered by this pattern.

FIG. 20 shows an example of a simple totalization processing pattern in the case of multi-answer. In the case of a single answer, the frequency itself of each option obtained as a result of the survey was displayed as the result of aggregation processing, but in the case of a multi answer, it is possible to duplicate multiple answers from the options and to select two levels. Totaling is required. That is, it is necessary to provide a frequency counter and integrate the frequencies selected for each option.

What is different from the case of the single answer is that it is meaningless to display the frequency thus obtained in% of the whole. In FIG. 20A, a frequency counter for each of the options is provided in association with the category number. In addition, it is shown in a picture that these are sorted and counted from the survey data.

In (b), the frequency counters such as "Company name", "Kasha", "Esha", "Isha", "Usha", "Asha", "Osha" are displayed in a square frame. The counted frequency is set to "9", "6", ... "3", "2" corresponding to the company name, and the maximum value of the counted frequency is "9" in the graph scale. The frequencies corresponding to company names are sorted by frequency. A bar graph corresponding to these frequencies is also displayed.

As in the case of the single answer above, the pattern is set so that the simple totalization process is performed by embedding the character string and the numerical value corresponding to the square frame, and the answer data to the question in the multi-answer format is simple. Can be aggregated. Naturally, the size of the summary table changes according to the number of options.

By patterning the simple totalization process as described above, the survey reply data can be first simply totalized for each question type, which makes it easy to predict the selection of variations of various analyzes of the survey results.

Although not shown here, other question types can be aggregated by using the question structure information, the data structure information, the survey data and the aggregation processing pattern.

FIG. 21 shows a flowchart (part 1) of the simple totalization process. Since the totaling process is different for each question type, FIG. 22 shows a flowchart in the case of a single answer (No. 2).

FIG. 21 shows the case where the simple totalization process corresponding to the given question number is performed. First, in step S211, the corresponding question structure information is extracted. Further, in step S212, the file name of the survey data in the data structure information corresponding to the same question number is taken out.
In step S213, this investigation data file is opened so that it can be input.

In step S214, the aggregation processing pattern storage unit 1 is associated with the question type obtained from the question structure information.
Start the aggregation processing program from 0. The aggregation processing program obtains information necessary for aggregation processing from the question structure information and performs simple aggregation.

FIG. 22 shows the operation of the simple tabulation program when the question type is single answer (SA). This corresponds to the step S214 in FIG. In step S221, the number of choices, that is, the number of items in the table and graph depending on the number of categories (C) specified, is an element that changes depending on the question. First, the area is secured corresponding to these.

In step S222, the parts corresponding to "sex" surrounded by a square frame in FIG. 19 are "male" and "female".
It is called a label of a name representative of the item name such as, and is set to the one designated by the option symbol (“”) of the question structure information.

In step S223, the contents of the category designation symbols are entered in the category column of the aggregation processing information by the number of categories designated in the question structure information. As a result, “male” and “female” in FIG. 19 are set.

In step S224, the survey data opened above is read, the frequency is counted for each category, and the frequency is entered in the frequency area. In step S225, the entered frequencies are rearranged in order of magnitude.

In step S226, the maximum frequency is set as the maximum scale of the graph. In step S227, the cumulative frequency, the ratio% to the whole, and the cumulative% are calculated from the entered frequencies and entered in the table.

Finally, in step S228, the simple totalization table and graph created are displayed in a visible form and output in a printed form.

[0104]

As is clear from the above description, according to the present invention, it is possible to realize labor saving from the preparation of the survey form to the data analysis and the efficiency improvement by reusing the survey know-how and the survey data. It has an industrial effect.

[Brief description of drawings]

FIG. 1 is a block diagram of the present invention.

FIG. 2 is an explanatory diagram of the configuration of the embodiment of the present invention (No. 1)

FIG. 3 is an explanatory diagram (part 2) of the configuration of the embodiment of the present invention.

FIG. 4 is an explanatory diagram (part 3) of the configuration of the embodiment of the present invention.

[Fig. 5] Example of question control statement

[Fig. 6] Example of survey form

[Fig. 7] Question structure information element by question type

FIG. 8 Question structure format table

[Fig. 9] Example of question structure information (No. 1)

FIG. 10: Example of question structure information (Part 2)

FIG. 11: Structure of Survey Form (Part 1)

FIG. 12: Structure of survey form pattern (Part 2)

FIG. 13 Structure of emphasized sentence pattern

FIG. 14: Flowchart of survey form preparation section

FIG. 15: Example of data structure information

FIG. 16 is a flowchart of the data structure information analysis unit.

[Figure 17] Example of survey data format

FIG. 18: Flowchart for survey data input

FIG. 19 is an example of a simple totalization processing pattern (No. 1)

FIG. 20 is an example of a simple tabulation processing pattern (part 2)

FIG. 21 is a flowchart of a simple totalization process (No. 1)

FIG. 22 is a flowchart of a simple totalization process (part 2)
[Aggregation in single answer (SA) format]

[Explanation of symbols]

 1 Question Structure Information Input Means 2 Question Structure Information Storage Units 3 Survey Form Creating Means 4 Survey Form Pattern Storage Units 5 Data Structure Information Analyzing Means 6 Data Structure Information Storage Units 7 Survey Data Input Means 8 Survey Data Storage Units 9 Aggregation Processing Means 10 Aggregation processing pattern storage

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Yasuyo Suzuki 1-6-1, Marunouchi, Chiyoda-ku, Tokyo Incorporated Fujitsu System Research Institute

Claims (6)

[Claims] 1. A questionnaire processing system for a questionnaire survey, in which a question control statement in which the structure of a question is described according to a prescribed notation according to a question type is input and divided into the question structure information storage section for each question. The question structure information input means to be stored, the questionnaire pattern storage unit that stores the questionnaire display format corresponding to the question type, the question structure information stored in the question structure information storage unit, and the questionnaire pattern storage unit A survey form processing system comprising: a survey form creating means for producing a survey form for each question based on a survey form display format. 2. A data structure information analysis means for extracting information on the structure of answer data for a question from the question structure information for each question stored in the question structure information storage unit and storing it in the data structure information storage unit, Based on the data structure stored in the data structure information storage unit, survey data input means for inputting response data to the survey form created above and storing it in the survey data storage unit,
The survey form processing system according to claim 1, further comprising: 3. An aggregation processing pattern storage unit storing an aggregation processing format corresponding to the question type, and survey data stored in the survey data storage unit are input,
A totaling processing unit for totaling for each question based on a totaling processing format stored in the totaling processing pattern storage unit corresponding to a question type of the question structure information storage unit is additionally provided. The survey form processing system according to claim 1 or claim 2. 4. A method for processing a questionnaire for a questionnaire survey, wherein a question control statement described based on a notation determined for each question type is input and stored, and the question structure information stored above and the question type stored in advance. A questionnaire processing method characterized in that a questionnaire is created for each question based on the questionnaire display format corresponding to. 5. A survey data storage unit for extracting information on the structure of answer data to a question from the question structure information and storing the data structure, and inputting the answer data for the created questionnaire based on the data structure. The survey form processing method according to claim 4, characterized in that 6. The survey data stored in the survey data storage unit is input, and aggregation is performed based on an aggregation processing format stored in advance corresponding to the question type of the question structure information. The survey form processing method according to claim 5.