Earticle

본 연구에서는 온라인 리뷰의 한 형태인 온라인 평점의 분산이 소비자 반응에 미치는 영향을 규명하고자 하였다. 구체적으로, 본 연구에서는 온라인 평점이 가장 빈번하게 쓰이는산업 중 하나인 영화 산업에 주목하였으며, 영화 평점의 분산이 크더라도 소비자가 항상 부정적으로 반응하지 않을 수 있음에 주목하였다. 즉, 영화 평점의 분포가 수렴하여 분산이 작은 조건과 분포가 발산하여 분산이 큰 조건에서 소비자가 보이는 반응은 소비자의 영화 관람동기에 따라 달라질 수 있다는 것이다. 본 연구에서는 구매동기와 관련된 선행 연구를 근간으로 영화 관람동기를 즐거움 추구 동기와 자기표현 동기로 분류하였으며, 즐거움 추구 동기를 가진 소비자들은 영화 평점이 수렴하는 조건(분산이 작은 조건)에서 영화에 대해 긍정적으로 반응하지만, 자기표현 동기를 가진 소비자들은 영화 평점이 발산하는 조건(분산이 큰 조건)에서 영화에 대해 긍정적으로 반응하게 됨을 규명하였다. 이제까지 온라인 리뷰 평점의 분산에 대한 소비자 반응을 다룬 연구들은 많이 이루어지지 않았는데, 본 연구는 영화 평점에 초점을 맞추어 영화 평점의 분산이 소비자 설득에 미치는 영향을 규명함으로써 온라인 리뷰에 대한 학문적 영역을 확장하였다. 또한 실무적으로는 온라인 마케팅과 관련된 실질적 지침을 제공하였다는데 의의가 있다.

This research aims to reveal how variance in online reviews influences consumer responses. Specifically, this research focuses on the movie industry, which uses online reviews to promote movies, and aims to reveal the phenomenon in which consumers may not show a negative response toward a movie even when the variance in online review scores for movie is large. This research suggests that motive for movie watching is a moderator, and argues that consumer responses toward a movie would differ depending on the motives for movie watching. This research divides motives for movie watching into motives of self-expression and motives of seeking pleasure, and shows that consumers with the motive of self-expression show positive responses toward movies with a large variance of online review scores, while consumers seeking pleasure show positive responses toward movies with a small variance in online review scores. This research is expected to provide practical implications for movie marketers, and expand the research scope of future online reviews.

This study examined the effects of various individual differences on children’s memory of a stressful experience. The participants for the current study were children (N=85) aged 4-9 years those who experienced a naturally occurring stressful experience from a dental procedure. There was overall negative relation between the level of stress and children’s memory performance. However, more interestingly, the results of this study provided some further evidence that several cognitive (i.e., receptive language ability and working memory capacity) and emotional (i.e., children's general anxiety condition, children’s self-report of pain and anxiety about the event) individual difference factors were associated with variations in children’s remembering across ages. The results suggest that the relation between stress and children’s memory might be impacted in part by children’s various individual characteristics. Furthermore, the findings are discussed in the applied context that based on the results clinical and legal professionals can tailor interviews to best meet children’s needs and capabilities, and create developmentally and individually sensitive guidelines for interviewing children in the legal system.

A crucial first step in the planning of any scientific experiment is to evaluate an appropriate sample size to permit sufficient statistical power to detect the desired effect. In this study, we investigated the optimal sample size of quantitative trait locus (QTL) linkage analysis for simple random sibship samples in pedigreed chicken populations, under the variance component framework implemented in the genetic power calculator program. Using the program, we could compute the statistical power required to achieve given sample sizes in variance component linkage analysis in random sibship data. For simplicity, an additive model was taken into account. Power calculations were performed to relate sample size to heritability attributable to a QTL. Under the various assumptions, comparative power curves indicated that the power to detect QTL with the variance component method is highly affected by a function of the effect size of QTL. Hence, more power can be achievable for QTL with a larger effect. In addition, a marked improvement in power could be obtained by increasing the sibship size. Thus, the use of chickens is advantageous regarding the sampling unit issue, since desirable sibship size can be easily obtained compared with other domestic species.

The objectives of this study were to quantify the combination values of the principal components and factors calculated using body measurements of Hanwoo(Korean Cattle) and estimate their heritabilities. The technique of multivariate analysis was used to reduce a large number of variables to a smaller number of new variables and characterize cattle according to body shape. The analyses were performed using 1,979 cattle at 12 months of age and 936 cattle at 24 months of age. The data for the analyses was obtained from progeny tests performed on Korean Cattle for 6 years from 2003 to 2008. The phenotypic correlations among these traits were estimated to range from 0.32 to 0.90 at 12 months of age and from 0.21 to 0.82 at 24 months of age. The first principal components(PC1s) indicated a weighed average of overall body measurements, accounting for 99.91% of the total variation for both periods of test. The two first PCs had positive coefficients for all body measurements. The major sources of PC, such as chest girth(CG), body length(BL), rump height(RH), and wither height(WH) were similar for both test periods. The heritabilities for PC1, the first factor score(FS1), and the second factor score(FS2) were estimated by multivariate REML method. The estimated heritabilities for PC1, FS1, and FS2 were 0.33, 0.38, and 0.40, respectively, at 12 months of age and 0.26, 0.76, and 0.58 at 24 months of age. Further studies are needed to determine whether the heritabilities of FS1 and FS2 at 24 months of age were overestimated.

We show that the highly volatile variance risk premium (VRP) can be theoretically and empirically reconciled with investor sentiment captured by temporary variation in risk aversion. In an effort to understand the poor predictive power of the VRP in non-U.S. markets, we propose a new investor sentiment index, the Variance Sentiment index (VSI), obtained from the trading behavior of individual investors. We show that the VSI predicts local return dynamics, in a similar way to what the VRP does in the US market. Moreover, the VSI does not lose its predictive power even in the presence of the global VRP.

Measurement error is one of main source of error in survey. It is generally defined as the difference between an observed value and an underlying true value. An observed value with error may be expressed as a function of the true value plus error term. In some cases, the measurement error variance may be also a function of the unknown true value. The error variance function can be rewritten as a function of true value multiplied by a scale factor. This research explore methods for estimation of the measurement error variance based on the data from complex sampling design. We consider the case in which the variance of mesurement error is a linear function of unknown true value, and the error variance scale factor is small. We applied our results to the U.S. Third National Health and Nutrition Examination Survey (the U.S. NHANES Ⅲ) data for empirical analyses, which has replicate measurements for relatively small subset of initial respondents's group.

This study aims to examine the influence of investor sentiment in the cryptocurrency market on the mean-variance relationship of Bitcoin. We find that Bitcoin expected excess return is positively related to conditional variance of Bitcoin return daily during the period of Extreme Fear but unrelated to conditional variance during the periods of Fear, Neutral, Greed, and Extreme Greed. Our findings are consistent with a distortion of a positive mean-variance tradeoff by investor sentiment. We also find that a negative relation between returns and contemporaneous innovations of conditional variance is only significant in the period of Extreme Fear. Moreover, the findings are robust, even after controlling for the impact of global factors determining Bitcoin price, and in Ethereum market, beyond Bitcoin market.

본 연구는 선형-선형 다중회귀모형을 통해 여과속도, 탈진압력, 탈진간격, 입구농도 등의 운전조건에서 벤츄리 설치가 압력손실에 미치는 영향을 예측하여 충격기류식 여과집진장치의 효율적인 운전조건과 경제적인 설계 기초자료로 활용하 고자 하였다. 여과면적 6 m2인 파이로트형 충격기류식 여과집진장치를 이용하여 제철소 coke분진으로 여과속도, 입구농 도, 탈진압력, 탈진간격 및 벤츄리 설치 유·무의 조건에 따라 여과포 압력손실을 측정하였으며, 여과포 압력손실(P)의 변화에 기여하는 여과속도(Vf), 입구농도(Ci), 탈진압력(PP), 탈진간격(Pi), 벤츄리 설치 유·무 등에 따른 영향력을 파악하 기 위해 SAS 9.4(SAS Institute, USA) 프로그램으로 통상적인 최소자승 추정법(Ordinary Least Square; OLS)인 선형-선 형의 다중회귀모형 함수를 이용하여 벤츄리가 압력손실에 미치는 영향을 예측하였다. 예측 결과를 종합해 보면 압력손실 최소화를 위해서 벤츄리를 설치할 경우 20.3% 효율 향상이 기대되며, 압력손실에 기여하는 순서는 여과속도(Vf) > 탈진 압력(Pp) > 입구농도(Ci) > 탈진간격(Pi)임을 알 수 있었다.

The aim of this study is to predict the pressure drop due to the presence or absence of Venturi under various operating conditions such as filtration velocity, pulse pressure and interval, and dust concentration using linear-linear regression model and use it as an efficient operating condition and economic data for a shock-type filter collection system. A pilot filtration collector with a filter area of 6 m2 was used, and the filtering pressure drop was measured according to filtration velocity, inlet concentration, pulse pressure, pulse interval, and the flow and absence conditions of the Venturi installation by a steel plant. The SAS 9.4 program (SAS Institute, USA) was used to predict and to determine the effects of filtration velocity (Vf), inlet concentration (Ci), pulse pressure (Pp) and interval (Pi), presence or absence of venturi, etc. and also analyze the changes in pressure drop according to the presence of Venturi using the linear-linear multiple regression model, and the least-squares estimation method (OLS). The results re shown below. In order to minimize the pressure drop, the improvement in efficiency is expected to be 20.3% when the Venturi is installed, and the order of contributing to the pressure drop is filtration rate> pulse pressure> inlet concentration> pulse interval.

We develop a discrete-time option pricing model using a variance-dependent pricing kernel of Christoffersen, Heston and Jacobs (2013) under an economic framework allowing for dynamic volatility and dynamic jump intensity. Using this model, we examine the role of the variance premium and jump risk premium in explaining S&P 500 index option returns. Our results stress the importance of the variance premium in explaining the stylized characteristics of index option returns, including short-term option returns, which are insufficiently explained by extant option pricing models. In particular, the variance premium can explain both 1-month holding period returns of 2-month maturity straddles, which are significantly negative, and call returns, which decrease according to moneyness. Even though the jump risk premium emphasized in the previous literature is able to well fit the option prices, it does not improve the explanatory power for the above two stylized option returns. The outperformance of the variance premium stems from its ability to capture the wedge between physical and risk-neutral volatilities.

To monitor all elements simultaneously of variance-covariance matrix Σ of several correlated quality characteristics under multivariate normal process Np(μ, Σ), multivariate exponentially weighted moving average (EWMA) chart and cumulative sum (CUSUM) chart are considered and compared. Numerical performances of the considered variable sampling interval (VSI) charts are evaluated using average run length (ARL), average time to signal (ATS), average number of switches (ANSW) to signal, and the probability of switch Pr(switch) between two sampling interval d1 and d2 where d1 < d2. For small or moderate changes of Σ, the performances of multivariate EWMA chart is approximately equivalent to that of multivariate CUSUM chart

본 연구는 1999년 1월부터 2004년 12월까지의 체세포점수의 검정일 자료를 이용하여 124,635개의 초산자료와 134,308개의 2산자료, 37,412개의 3산자료, 41787개의 4산자료, 37412개의 5산 자료를 이용하였다. 분석에 사용된 방법은 체세포 점수에 영향을 미치는 분만연도, 연령, 비유단계, 산차, 계절의 효과를 추정하기 위하여 GLM을 이용하여 최소자승법으로 분석하였으며, 검정일 모형을 이용한 분산성분 추정은 EM-REML 분석방법을 전산 프로그램한 REMLF90을 이용하였다. 각 산차별로 연령에 대한 효과는 연령이 낮은 군에서 체세포 점수가 낮게 나타났으며, 연령이 높은 군에서는 다소 높게 나타났다. 비유단계별 효과는 1산과 2산에서는 비유초기에서 체세포 점수가 낮게 나타났으며 비유말기에 3.151, 3.696로 높은 체세포 점수를 나타났으나, 3산, 4산, 5산에서는 비유중기에 높은 체세포 점수를 나타내었으며, 대체로 피크기에 체세포 점수가 높게 나타나는데 4산과 5산에서는 비유말기에 체세포 점수가 낮게 나타났다. 분만계절별 환경효과는 1산～5산 모두 대체로 9～11월에 체세포 점수가 낮게 나타났으며, 대체로 유량이 낮게 추정되는 6～8월 사이에는 체세포 점수가 대체로 높게 나타났다. 각 산차별 유전력은 1산에서 5산까지 각각 0.05, 0.09, 0.10, 0.05, 0.05를 나타냈으며, 유전분산값은 비유초기의 경우 2, 3, 5산에서 높게 나타났으며, 1산과 4산의 경우는 대체로 낮게 추정되었다.

This study utilized test day of somatic cell score data of dairy cattle from 2000 to 2004. The number of data used were 124,635 of first parity, 134,308 of second parity, 77,862 of third parity, 41,787 of forth parity and 37,412 of fifth parity. The data was analyzed by least square mean method using GLM to estimate the effects of calving year, age, lactation stage, parity and season on somatic cell score. Variance component estimation using test day model was determined by using expectation maximization algorithm- restricted maximum likelihood (EM-REML) analysis method. In each parity, somatic cell score was low for younger group and was relatively high in older groups. Likewise, for lactation stage, the score was low in early-lactation and high in late-lactation in first parity and second parity. Nevertheless, for the third, fourth and fifth parity, however, high somatic cell score was observed in mid-lactation. Generally, the score was high in the peak. Although in fourth and fifth parity, the score was low in late-lactation. Environmental effect of season, somatic cell score was generally low from September to November for all parities. The score was high between June and August when the milk production is usually low. The heritability in each parity were 0.05, 0.09, 0.10, 0.05 and 0.05 for parity 1, 2, 3, 4, 5, respectively. Genetic variance value was estimated to be high in second, third and fifth parity in early-lactation and to be low in first and forth parity.

본 연구에서는 내담아동이 지각하는 치료관계에 영향을 미치는 치료자변인을 밝히고자 학령기 아동과 치료자 총 109명을 대상으로 하여 연구하였다. 척도는 신경일(1994)의 공감 능력 검사지와 방은령, 최명선(2004)의 치료관계 척도를 사용하였고 자료 분석은 Cronbach' α 계수, t-test, ANOVA, 중다회귀분석을 실시하였다. 연구결과는 첫째, 내담아동들은 많은 상담사례를 한 치료자에게 치료적 협력관계를 더 강하게 형성하기도 하지만 부정적 감정도 더 강하게 지니고 있었다. 그러나 치료자의 총상담회기는 내담아동이 지각하는 치료적 협력관계는 치료자의 총상담회기가 많을수록 높아지지만 감정관계는 치료회기 집단에 따라 다른 양상을 보였다. 둘째, 놀이치료자가 슈퍼비젼 받은 기간이 길고, 횟수가 많을수록, 내담아동은 치료적 협력관계를 더 높게 지각하였다. 또한 개인분석을 받은 경험이 없는 집단보다 있는 치료자들에게, 개인분석 기간이 1-2년 미만이고 횟수가 51-100미만일 때 내담아동들은 치료적 협력관계를 가장 잘 형성한다는 것을 알 수 있었다. 그리고 치료자의 슈퍼비젼 횟수가 100회 이상이거나 1-10회일 때 내담아동들은 치료자에 대해 긍정적인 감정을 가장 강하게 형성하고 있었고, 치료자의 개인분석 기간이 2-3년일 때 부정적 감정관계가 가장 강하게 지각되었다. 셋째, 놀이치료자가 공감적 관심과 정서적 공명을 높게 지닐수록 내담아동은 치료자에게 보다 더 치료적으로 협력적이었다.

The purpose of this study is to examine which relation between the play therapist's Variance and a therapeutic relation perceived by Client Children in the playtherapy. The subjects of this study were 109 children older than 8 years, their therapist. The korean version of the therapeutic relationship scaled for children The collected data were analyzed by frequency, percentage, Cronbach's α, t-test, ANOVA, multiple Regression. First, Playtherapists have more negative emotional relation by Client Children when they are unmarried. and have less therapeutic helping relationship when they work in childcounseling Center in Hospital. Second, it was shown that the more a play therapist has the number of cases, the more therapeutic helping relationship and. higher negative emotional relationship to therapists. It was shown that a play therapist has more total Sessions, the higher therapeutic helping relationship. But the emotional relationship was not show the same Trend. Third, it was shown that the more a play therapist has supervision experience, the higher therapeutic helping relationship. and as far as having the personal counseling experience more than 51-100 times or 1-2 years, it was shown that the more a play therapist has personal counseling experience, the higher therapeutic helping relationship. and as far as having the supervision experience more than 100 times or 1-10 times, it was shown that the more a play therapist has supervision experience, the higher positive emotional relationship. Examined with subfactors of empathy of playtherapists and therapeutic helping relationship felt by client children, it is revealed that playtherapists have more empathic concern and emotional resonance, client children feel more therapeutic helping relationship to therapists.

일상생활 속에서 다양한 밀도와 크기를 가진 사물들의 개수를 추정하는 것은 인간이 가진 중요한 능력이다. 그러나 기존의 개수 추정 연구들 대부분은 자극의 크기 변산을 주요 변인으로 고려하지 않아 다양한 환경 속에서 개수를 추정할 수 있는 인간의 능력을 한정적으로밖에 보여주지 못했다(Burr & Ross, 2008a; Durgin, 1995). 본 논문에서는 개수 추정에 중요한 두 변인인 밀도와 크기 변산이 개수 추정에 미치는 영향을 연구하였다. 연구 결과, 밀도와 크기 변산은 상호 작용 없이 개수 추정에 독립적인 영향을 미쳤다. 구체적으로, 밀도와 크기의 다양성이 커질수록 개수는 과소 추정되었다. 또한 개수 추정이 밀도를 통해 간접적으로 추정된다는 기존 연구 결과와는 달리, 본 연구에서 개수는 밀도에 상관없이 직접적으로 추정되는 것으로 나타났다. 밀도를 통하지 않고 직접적인 개수 추정이 가능했던 것은 자극의 크기가 다양하여 자극들이 개별적으로 잘 구분되었기 때문인 것으로 보인다.

In our daily life, it is important to estimate the approximate number of objects containing various sizes and appearing in different densities. However, previous studies mostly used dot stimuli with uniform sizes (Burr & Ross, 2008a; Durgin, 1995). In this study, we investigated how density and size variance influenced numerosity estimation. We found that numerosity was underestimated when density and size variance increased. There was no significant interaction between the density and the size variance. Also, we observed that numerosity was directly estimated by the number mechanism itself across all the range of density, different from the results of previous research. We suggest that this different result is due to various sizes used in this study, which helps the visual system to individuate each object to count.

최근 상업용 통계 패키지의 발달에 힘입어 다변인 통계분석에 기초한 연구들이 양산되고 있다. 다변인 분석은 단변인 통계에 비해 복잡한 연구문제를 탐색할 수 있는 장점이 있지만, 통계분석의 복잡성으로 인하여 절차상 오류들이 더러 발견된다. 더욱이, 다변인 통계분석의 수학적 기초가 복잡하여 그 원리를 이해하기 어렵고, 분석방법 및 결과 해석이 난해할 수 있으며, 다양한 분석법들이 상반된 결과를 낳아서 결과해석이 곤란해질 수 있다. 따라서 본 연구에서는 다변인분석 적용상의 가능한 오류와 주의점들을 기술하여 이러한 오용을 줄일 수 있는 방안을 제시하고자 하였다.

Due to the recent developments in commercial statistical packages, vast amount of research articles based on multivariate statistics have been published. While the multivariate statistics have advantages over the univariate statistics in dealing with more complicated research hypothesis, abuses and misuses in it's application have been frequently noticed in many researches, due to the underlying complexities. Specifically, it is difficult to understand the basic principle due to the mathematical complexities underlying in it, henceforth making the interpretation of the analysis result be difficult. Even worse, different choice of the analysis methods sometimes result in contradicting results, making the interpretation implausible. This study tried to examine the cases of misuses in the application of multivariate statistics, and to identify the problems. Finally, several precautions for reducing the incidents was proposed.

정서조절의 실패, 즉 정서 불안정성은 우울을 설명하는 핵심적인 기제 중 하나이다. 선행 연구에서 정서 불안정성은 주로 정서 변산성과 정서변화 저항성으로 구분되는데, 전자는 분산(variance)을 통해, 후자는 자기상관(autocorrelation)을 통해 산출된다. 본 연구에서는 기존의 지표들이 우울한 개인의 정서 불안정성을 제대로 반영하고 있는지에 대한 질문에 답하기 위해 경험표집법(experience sample method)을 통해 수집된 자료를 분석하여 해당 지표를 산출하고, 나아가 정서 불안정성을 반영하는 또 다른 지표인 계차제곱평균(Mean Square Successive Difference, MSSD)을 추가적으로 산출하여 이를 비교하는 것을 주된 목적으로 하였다. 또한 기존의 연구에서 주로 강조된 정서의 유인가(valence)에 더하여 각성가(arousal)의 차원까지 고려하여 우울과 관련된 정서 불안정성의 특징을 보다 포괄적으로 살펴보고자 하였다. 이를 위해 우울 성향을 보이는 집단(25명)과 그렇지 않은 집단(24명)의 참가자에게 긍정-각성(예: 행복한, 유쾌한), 긍정-비각성(예: 평온한, 편안한), 부정-각성(예: 짜증난, 불안한), 부정-비각성(예: 우울한, 피곤한)의 차원에서 자신의 기분을 하루에 5번씩 7일 연속으로 7점 척도 상에 보고하도록 하였다. 집단 간 차이 검증의 결과는 다음과 같다. 첫째, 분산을 통해 측정된 정서 변산성의 경우, 우울 집단이 비우울 집단보다 유인 및 각성 변산성과 질적 변산성에서는 높은 수준을 보였으나 강도 변산성에서는 차이를 보이지 않았다. 둘째, 자기상관을 통해 측정된 정서변화 저항성의 경우, 유인 및 각성 차원에서 모두 차이를 보이지 않았다. 셋째, 계차제곱평균을 통해 측정된 불안정성의경우, 우울 집단은 비우울 집단보다 유인 및 각성 차원에서 모두 높은 수준을 보였다. 본 연구는 우울증의 주요 특징인 정서 불안정성을 살펴보기 위해 유인가와 각성가를 모두 고려하여 분산, 자기상관 및 계차제곱평균을 비교하고 우울과 관련된 정서조절의 특징을 통합적으로 분석하였다는 데에 의의가 있다. 마지막으로 연구의 제한점 및 후속 연구의 방향을 제안하였다.

Depressive disorders are known to be closely related to emotional dysregulation and therefore, emotional instability. Previous studies suggested that emotional instability can be characterized by two components, emotional variability and emotional inertia, the indice of which are emotional variability and emotional inertia, respectively, but there have been few studies which comprehensively examined emotional instability of depressed individuals. The current study, by utilizing experience sampling method, aimed to compare variance and autocorrelation with the Mean Square Successive Difference, a newly introduced index of emotional instability reflecting both variability and inertia of emotion. In addition, this study aimed to examine the dynamics of emotional instability with respect to activation dimension as well as valence dimension of emotion. For this purpose, 25 participants with depression and 24 non-depressed participants were asked to rate their positive-activated (happy, pleasant), positive-deactivated (calm, comfortable), negative-activated (irritated, anxious), and negative-deactivated (depressed, tired) emotional states, five times a day for seven consecutive days. The results of the independent sample t-tests were as follows: First, participants with depression displayed higher valence variability, activation variability, and spin compared with non-depressed individuals, while no significant group difference in pulse was found. Second, there was no significant group difference in valence and activation of emotional inertia. Third, the depressed group showed significantly higher scores in valence and activation of mean successive difference, compared with the non-depressed group. The current study is significant with respect to the synthetically analyzed dynamics of emotional instability by comparing three major indexes including variance, autocorrelation, and mean square successive difference, as well as by considering two dimensions of emotion.

본 연구는 두 집단 분포의 등분산을 가정하는 JZS(Jeffreys-Zellner-Siow) 사전분포 설정과 이분산을 허용하는 BFGC(Girón-del Castillo) 사전분포 설정을 이용한 베이지안 독립표본 t 검정의 타당성을 비교하였다. 특히 표본크기 비율과 분산 비율의 조합이 베이즈 인자의 산출에 미치는 영향을 시뮬레이션 연구를 통해 확인하였다. 시뮬레이션은 분산 비율(1:1, 2:1), 표본크기 비율(1:1, 2:3, 3:2), 표준화 효과 크기(0, 0.2, 0.5, 0.8), 총 표본크기(50, 100, 200)를 교차한 조건에서 수행되었으며, 각 조건당 500회의 반복을 통해 생성된 총 30,000개의 데이터 세트가 사용되었다. 분석 결과 두 사전분포 설정이 산출한 베이즈 인자는 이분산 조건에서 분산이 큰 집단의 표본크기의 상대적 크기에 따라 상반된 양상이 확인되었다. JZS 설정은 분산이 큰 집단의 표본크기가 상대적으로 큰 경우 참인 가설에 대한 지지가 약화되는 반면, BFGC 설정은 분산이 큰 집단의 표본크기가 상대적으로 작은 경우 유사한 양상을 보였다. 이러한 결과는 베이지안 t 검정에서 등분산 가정이 표본크기 및 분산 비율과 상호작용하여 베이즈 인자의 크기에 체계적인 영향을 미친다는 점을 보여준다. 또한 이분산이 의심되는 실제 연구 상황에서 등분산 가정에 기반한 JZS 모형보다 이분산을 허용하는 BFGC 모형을 사용하는 것이 더 나은 베이지안 추론을 제공할 수 있음을 시사한다.

Heterogeneity of variance is a persistent concern in independent-samples t tests, raising questions about the robustness of Bayesian hypothesis testing when the equal-variance assumption is violated. The Jeffreys-Zellner-Siow (JZS) prior, commonly used as the default in Bayesian t tests, inherently assumes homoscedasticity. The present study examines the implications of this assumption by comparing a homoscedastic Bayesian t test based on the JZS prior with a heteroscedastic alternative that allows group-specific variances, the Giron-del Castillo (BFGC) model. An extensive simulation study was conducted to investigate how Bayes factors behave across varying combinations of variance ratios, sample size ratios, standardized effect sizes, and total sample sizes. Particular attention was given to conditions in which sample size imbalance interacted with variance heterogeneity. The results showed that the two models exhibit qualitatively different patterns of evidence accumulation under heteroscedasticity. Specifically, the JZS-based Bayes factor tended to provide weaker support for the true hypothesis when the group with the larger variance also had the larger sample size, whereas the BFGC-based Bayes factor showed the opposite pattern, yielding weaker support when the larger-variance group had the smaller sample size. These findings highlight that variance assumptions in Bayesian t tests can systematically influence the interpretation of Bayes factors, especially in the presence of sample size imbalance. When heteroscedasticity is plausible, adopting a heteroscedastic Bayesian model such as BFGC may therefore lead to more reliable Bayesian inference than reliance on the default JZS specification.

본 논문은 Campbell 등(1988, 1991, 1993a, 1993b)이 VAR접근법을 이용하여 도출한 주가수익률 변동성 분해 실증결과의 신뢰성에 대해 의문을 제기하고, 대안모형을 통해 재평가해 보았다. VAR접근법의 계량경제학적 문제점은 다음과 같다. 첫째, Campbell 등이 사용한 VAR 모형의 추정계수들 가운데 통계적으로 유의하지 않은 것이 다수 존재하며, 따라서 이를 토대로 한 주가수익률 변동성 분해도 정확도가 떨어질 수 밖에 없다. 둘째, VAR 모형을 구성하고 있는 주가수익률 예측방정식의 경우 배당수익률 변수에 대한 계수 추정에 있어 상향편기가 발생할 가능성이 크며, 그 결과 VAR접근법에 의한 변동성 분해는 미래할인율 요인을 과대 측정하는 경향을 보이게 된다. 특히 한국전쟁 이후 이와 같은 모습이 더욱 크게 나타났으며, 배당수익률 계수의 상향편기를 조정한 후 설명력을 재 측정해 본 결과 미래할인율 요인의 설명력이 크게 감소하였음을 확인할 수 있었다. 마지막으로 본 논문은 VAR접근법과 대안모형인 회귀접근법에 의한 주가수익률 변동성 분해를 상호 비교해 보았는데, 회귀분석접근법을 사용할 경우 한국전쟁 전후 모든 기간에 걸쳐 배당성장률 변수의 설명력이 큰 반면 미래할인율 요인이 설명하는 부분은 상대적으로 작게 나타나 VAR접근법에서의 실증결과와 합치하지 않았다

This paper casts doubt on the variance decomposition of the stock returns by the VAR approach from the econometric perspective. The econometric problems of the VAR approach are as follows: First, the parameters of the multivariate VAR model are prone to be estimated imprecisely so that the variance decomposition yielded by the VAR approach is not statistically significant. Secondly, the contribution of the future return component is likely to be overestimated due to an upward bias in estimating the coefficient of the dividend-price ratio variable in the stock return forecasting equation. In fact, we observe upward estimation biases in both pre-Korean war and post-Korean war periods, and the bigger bias in the post- Korean war periods. When we reestimate the contribution of the future return component, it is no longer as large even in the postwar periods. Finally, we compare the variance decomposition by the VAR approach with that by the regression approach. When the regression approach is used, we find a high explanatory power of the future dividend growth variables in both pre-war and post-war periods, and do not find a support for the results yielded by the VAR approach.