객관식 일변도 평가 방식의 문제점과 그 개선 방향 (토론)

安昌圭 한국교육과정학회 1988 교육과정연구 Vol.7 No.-

어린이 말에 있어서의 音素의 發達

安昌圭 新羅大學校 1977 論文集 Vol.5 No.-

This Paper attempts to investigate how Korean infants develop their phonemic system in the process of acquiring their native speech. For this purpose I have selected 188 infantile word which are meaningful to their life and contain most of the Korean phonemes. These words were tested on 29 normal infants (102 males and 107 females), all randomly selected in the Busan area, age between 7 months and 24 months, in order to see to what degree they could be pronounced distinctly. 154 students of Busan Women's College helped me with this investigation by collecting other infant vocabulary. The following conclusions were obtained from the stastistical analysis of the test; 1. The first word Korean infants use is 'eumma'; the next one, 'appa'. The phonemes /m/, /p/, and /a/ are discerned in the babies, whose age is 10 months and over. 2. The consonant phonemes to be recognized in the babies, age 10 to 24 months, are /m/, /p/, ??, /t/, /n/, /m/, /s/, /k/, /h/, ??, ??, ??, /c/ and /l/ in their order of eccurrence the vocal phonemes are /a/, /a/, /u/, /i/, /o/, /j/, /e/, and /w/ in their order of occurrence 3. Phonemic development in the Korean infant proceeds according to Jacobson's Law of Irreversible Solidarity. 4. Phonemic development in the Korean infant generally conforms to the developmental principles of other languages, including English, with a few exceptions, which strongly suggests the existence of 'language universal' concerning the acquisition of phonemes.

釜山市 就學前 兒童의 家庭敎育實態에 관한 調査硏究

安昌圭 新羅大學校 1979 論文集 Vol.7 No.-

Preschool education has been emphasixed in this country by many educational researchers and educators. Their ideas are based on the relative importance of early experience in child development and they try to analyze and control environmental experiences in early childhood including first year olds, by extracting main variables from the. This study is converned with analyzing home environment which gives early developmental experience and investigating actual parental practices on home education to facilitate child development. It could be possible to facilitate early childhood education effectively by recognizing actual state of educational home environment before emphasizing and practicing it. For the study, 268 mothers who had chilidren of 3 through 6 years of age and who had no regular job were randomly selected and individually interviewed by a structured questionnaire on educational home environment. The following conclusions could be extracted from the results obtained. 1) Home envirenment of early childhood in Busan families was inadequate for fachilitating function was relatively weaker. 2) The influential person to facilitate child development ws its mother while fathers' function was relatively weaker. 3) Most of the motehrs who had preschool children were suffering from some develpmental behaviro problems of their children. 4) Most of the motehrs' child-rearing attitudes were not based on psychological knowledge of facilitating heir children's development but on the natural and maturational growth in child development. 5) In spite of the motehrs'recognition on the relative importance of home environement in early childhood, most of the mothers had difficulty in deficiency of practical knowledge of child education, economic problem, deficiency of time to teach their children and inadequate facilities of nursery school or child caring centers near their home. 6) Most of the mothers were suffering form deficiency of information on child education and child-rearing practices. 7) the above mentioned problems were serious in the mothers of lower econmic income. Conclusively, it is urgently needed to perform adult education for mothers and to establish more schools or child caring centers for mothers in lower econmic income in order to enrich educational home environment for faciitating early childbood development

因果的 屬性知覺에 미치는 情報의 效果

安昌圭 新羅大學校 1976 論文集 Vol.4 No.-

An individual assigns causals to various responses he makes or observes and the consequenres of his resulting beliefs about causality. Kelley's conceptualization details the cognitive processes that might be engaged in by the individual as he interprets a given response in the context of the information which is characterized by high or low consensus, high or low distinctiveness, and high or low consistency. According to his prediction, external attribution is made when evidence exists as to the censensus distinctiveness, and consistency of the appropriate effects. The present study was to investigate the following aspects of Kelley's model; How and what causal attributions are farilitated by various combinations of consensus distinctiveness, and consistency information ? And what is the relative importance ofthese variables ? In an attempt to answer these questions, consensus, distinctiveness, and consistency information were manipulated orthogonally. The subjects (63 freshmen in womenln's college) were told of the occurrence of a response by another person. They were then given one of the eight possible combinations of consensus, distinctiveness, and consistency information regarding that response, And they were asked to indicate whether they thought something about the actor, something about the stimulus entity, somenthing about the circumstance, or combination of these three factors probably caused the response to occur. To provide a satisfactory means of answering the questions above, a group of control items subiects were given no information about the responses to be attributed. The results obtained were as follows : 1) Under the condition that no information was given to them, the individual lends to respond more often to the actor than to the stimulus or others in his causal inference. 2) As predicted HHH information patterns produced significantly more stimulus attirbution than did no or any other information patterns, LLH information patterns showed more actor attribution, and LHL petterils gave more circumstance attribution. 3) The consensus and distinctiveness effects were somewhat qualified by interactions with consistency information, and consensus and distinctiveness information had a substantial effect on person and stimulus attribution. 4) The relative importance of the three information variables varied so newhat a non: attirbution loci.

교육 및 심리검사에 있어서 문항반응이론의 성격과 그 적용성

안창규 한국교육평가학회 1987 교육평가연구 Vol.2 No.2

인지치료이론의 원리

안창규 부산대학교 학생생활연구소 1995 硏究報 Vol.30 No.1

면접고사의 합리적 방법과 문제에 대한 토론

안창규 한국교육평가학회 1989 교육평가연구 Vol.3 No.2

非同等統制集團設計에서 平均을 向한 回歸現象

安昌圭 부산대학교 사회과학대학 1984 社會科學論叢 Vol.3 No.1

The regression toward the mean have been emphasized as one of bias factors on estimating true experimental effects from nonequivalent control group designs, in which selection differences between the experimental and the control groups are inevitably introduced at the beginning of an experiment. However, a clear and full explanation of the logic behind the statistical models available to such nonequivalent control group designs has not been discussed according to the regression toward the mean effect. In this article, it is discussed about the nature and causes of the regression toward the mean phenomenon, the problems of the regression artifacts in estimating true experimental treatment by several statistical models in the nonequivalent control group designs with and without controlled selection based on a measured and quantified dimension of a pretest. The following several facts are conclusively clarified by the analysis of the regression toward the mean effect. First, the regression toward the mean phenomenon is caused by less than a unity of the pretest-posttest correlation atributable to the measurement errors of the tests, the differential growth rates of individuals, and/or changing structure of behavior. Second, as a deterious factor in estimating true experimental treatment effects in the nonequivalent control group designs, the regression artifacts are differential growths between the groups with initial selection differences and subsequent differential regression toward the mean effects. Third, such statistical models as ANCOVA model, and nonparall regression model available to the nonequivalent control group design with controlled selection can statistically control over the differential regression toward the mean effect between the groups. Accordingly, if possible in a situation inevitably encountered to the nonequivalent control group design, it is highly recommended to assign subjects to the experimental and the control groups on the basis of a measured and explicitly quantified dimension of the pretest.

골퍼의 기술수준별 골프클럽구매행동 및 브랜드태도에 관한 연구

안창규,백송원 한국스포츠리서치 2006 한국 스포츠 리서치 Vol.17 No.3

VRML을 이용한 Web기반의 가상공간 저작도구

안창규,이강우,이기동 嶺南大學校 工業技術硏究所 2004 工業技術硏究所論文集 Vol.32 No.1

In this paper, we introduce the design and implementation technique for WVRAT(Web Virtual Reality Authoring Tool) which is an easy tool for describing virtual space. The problems of the prevalent software such as, complexity of the GUI and high cost, prevent end users from widely using of VRML. Thus we propose an authoring tool for user to easily implement his own virtual space only combining the virtual objects in the network and the real data of individual server.