To ensure test security and fairness, alternative forms of the same test are administered in practice. However, alternative forms of the same test generally do not have the same test difficulty level, even though alternative test forms are designed to be as parallel as possible. Equating adjusts for differences in difficulties among forms of the test. Six traditional equating methods are considered in this study: equipercentile equating without smoothing, equipercentile equating with pre-smoothing and post-smoothing, IRT true-score and observed-score equatings, and kernel equating. A common feature of all of the traditional procedures is that the end result of equating is a single transformation (or conversion table) that is used for all examinees who take the same test. Van der Linden has proposed conditional equipercentile (or local) equating (CEE) to reduce the error of equating contained in the traditional equating procedures by introducing individual level equating. Van der Linden's CEE is conceptually closest to IRT-T in that CEE is with respect to a type of true score (θ, or proficiency), but it shares similarities with to IRT-O in that CEE uses an estimated observed score distribution for each individual θ to equate scores using equipercentile equating. No real-data study has yet compared van der Linden's CEE with each of the traditional equating procedures. Indeed, even for the traditional procedures, no study has compared all six of them simultaneously. In addition to van der Linden's CEE, two additional variations of CEE are considered: CEE using maximum likelihood (CEE-MLE) and CEE using the true characteristic curve (CEE-TCC). The focus of this study is on comparing results from CEE vis-à-vis the traditional procedures, as opposed to answering a “best-procedure”question, which would require a common conception of “true”equating. Although the results of the traditional equating methods are quite similar, the kernel equating method and equipercentile equating with log-linear presmoothing generally show better fit to the respective original form statistical moments under various data conditions. Although IRT-T and IRT-O usually are found to be least favorable under all circumstance in terms of statistical moments, the equated raw score difference distribution illustrates more stable performance than traditional equating methods. It was found here that the number of examinees having a particular score point does not influence results for CEE as much as it does for traditional equatings. CEE-EAP and CEE-MLE are very similar to one another and the equated score difference distributions are similar to those of IRT-O. CEE-TCC involves a part of the IRT-T procedure. Hence, CEE-TCC behaves somewhat similar to IRT-T. Although CEE results are less desirable in terms of maintaining statistical moments, the equated score differences are more consistent and stable than for the traditional equating methods.

College students often enter college academically unprepared, as evidenced by low high school cumulative GPAs or poor SAT scores. In response to this problem, administrators at a 4-year university in the Mid-Atlantic region of the Unites States implemented an intensive, semester-long program to introduce and acclimate conditionally admitted students to the rigors of collegiate life. The purpose of this study was to understand how to assist students in moving from Year 1 to full admission and beyond. In accordance with Bandura's reciprocal causation of social cognitive theory model, the research questions centered on conditionally admitted students' descriptions of their experiences with intensive, semester-long program participation. The qualitative case study used data collected from 10 semistructured interviews with conditional admission program student participants. Data analysis consisted of initial coding, axial coding, and iterative recategorization to identify the key findings. Among the findings were that the study site lacked strong faculty-student engagement and that students had mixed feelings regarding the seminar course being helpful. However, they found the university environment conducive to learning, leading them to stay. A white paper provided potential solutions to administrators, including increased faculty-student engagement and more meaningful required seminars for first-year conditionally admitted students. This study and the subsequent project may create positive social change by expanding degree achievement for underprepared, conditionally admitted college students, which thus increases opportunities for upward social mobility.

Von Willebrand Disease (VWD) is the most common bleeding disorder. In addition to known major genes, genetic modifiers, such as ABO blood group, affect quantitative outcome measures for VWD. The data consist of an 854-member Amish pedigree with established linkage of VWD to a locus within the Von Willebrand Factor (VWF) gene on chromosome 12. The DNA sequence of the causative mutation is known. Phenotypic information and genotypic data consisting of VWF mutation status and a genome screen of markers are available for 385 pedigree members. Genetic modifiers of the VWF mutation are investigated using known and new conditional linkage methods that search for modifier genes of a major gene with known mutation. The MCMC-based program LOKI was used to conduct multipoint linkage analysis of VWD outcome measures while controlling for the VWF mutation. Adjustment for the mutation did not eliminate the linkage signal on chromosome 12 in the same location as the VWF mutation. Evidence for QTLs was also found on six other chromosomes. Smod, a score statistic that detects evidence of a genetic modifier conditional on linkage to a major gene, was developed for sib pair data. To limit the modifier gene main effect, Smod was developed so that variance due to the modifier locus is bounded above by the variance of the interaction between major gene and modifier gene. The performance of Smod was compared to other published score statistics. Power to detect linkage to the modifier locus depended on major gene and modifier gene risk allele frequencies, relative contribution of the major gene main effect to the interaction effect, and the upper bound on the modifier gene main effect. The Amish pedigree was broken up into sib pair data and analyzed using Smod and other score statistics. Using these statistics, the strongest evidence for QTLs for VWD was also found on chromosome 12 in the region of the VWF mutation. Combined with the LOKI results, further analysis will help determine if intragenic modification is occurring or if linkage disequilibrium between the mutation and analyzed markers is driving results.

In recent years, there has been an increasing interest in the study of human operant behavior. One area of study reflecting this interest is the study of the formation of equivalent classes of stimuli by human subjects. The focus of the present research was the study of the conditions under which classes of equivalent stimuli can be inferred to be under conditional control. In Experiment 1-A, three college students were trained to respond to a balanced five-term contingency via a visual-visual simultaneous matching-to-sample task with two choices of comparison stimuli. Probe tests showed that subjects' behavior could be described as being controlled by positive and negative stimulus relations. When the second-order stimulus was removed during subsequent probes, none of the three subjects demonstrated strong correct responses to the four-term unit relations. Also, none of the three subjects demonstrated the expected transitive relations when the second-order (five-term) stimulus was removed. In Experiment 1-B--with the same three subjects--explicit training of the four-term unit relations showed the expected transitive relations in the absence of the second-order stimulus. In Experiments 2 through 5--using a matching-to-sample task similar to that used in Experiments 1-A and 1-B--five subjects were trained to respond to comparison stimuli C and E in the presence of sample A and second-order stimulus X and to comparison stimuli Band Fin the presence of sample D and second-order stimulus X. Likewise, the subjects were trained to respond to comparison stimuli Band Fin the presence of sample A and second-order stimulus Y and to comparison stimuli C and E in the presence of sample D and second-order stimulus Y. Probe tests for transitive relations showed that four of the five subjects eventually demonstrated four three-member classes of equivalent stimuli that functioned separately under the control of the second-order stimuli. The four subjects demonstrating the classes of equivalent stimuli either a) demonstrated the transitive relations immediately orb) demonstrated the transitive relations after explict retraining of the underlying four-term unit relations. The results of all experiments together indicated that the composition of classes of equivalent stimuli can be conditionally controlled by either a) removing the second-order stimulus orb) training subjects to respond to classes of equivalent stimuli under the control of other explicit visual second-order stimuli. The results are discussed in terms of verbal behavior, emergent behavior, and conceptual development.

Designing generalized data-driven distance measures for both ordered and unordered set data is the core focus of the proposed work. An ordered set is a set where time-linear property is maintained when distance between pair of temporal segments. One application in the ordered set is the human gesture analysis from RGBD data. Human gestures are fast becoming the natural form of human computer interaction. This serves as a motivation to modeling, analyzing, and recognition of gestures. The large number of gesture categories such as sign language, traffic signals, everyday actions and also subtle cultural variations in gesture classes makes gesture recognition a challenging problem. As part of generalization, an algorithm is proposed as part of an overlap speech detection application for unordered set. Any gesture recognition task involves comparing an incoming or a query gesture against a training set of gestures. Having one or few samples deters any class statistic learning approaches to classification, as the full range of variation is not covered. Due to the large variability in gesture classes, temporally segmenting individual gestures also becomes hard. A matching algorithm in such scenarios needs to be able to handle single sample classes and have the ability to label multiple gestures without temporal segmentation. Each gesture sequence is considered as a class and each class is a data point on an input space. A pair-wise distances pattern between to gesture frame sequences conditioned on a third (anchor) sequence is considered and is referred to as warp vectors. Such a process is defined as conditional distances. At the algorithmic core we have two dynamic time warping processes, one to compute the warp vectors with the anchor sequences and the other to compare these warp vectors. We show that having class dependent distance function can disambiguate classification process where the samples of classes are close to each other. Given a situation where the model base is large (number of classes is also large); the disadvantage of such a distance would be the computational cost. A distributed version combined with sub-sampling anchor gestures is proposed as speedup strategy. In order to label multiple connected gestures in query we use a simultaneous segmentation and recognition matching algorithm called level building algorithm. We use the dynamic programming implementation of the level building algorithm. The core of this algorithm depends on a distance function that compares two gesture sequences. We propose that, we replace this distance function, with the proposed distances. Hence, this version of level building is called as conditional level building (clb). We present results on a large dataset of 8000 RGBD sequences spanning over 200 gesture classes, extracted from the ChaLearn Gesture Challenge dataset. The result is that there is significant improvement over the underlying distance used to compute conditional distance when compared to conditional distance. As an application of unordered set and non-visual data, overlap speech segment detection algorithm is proposed. Speech recognition systems have a vast variety of application, but fail when there is overlap speech involved. This is especially true in a meeting-room setting. The ability to recognize speaker and localize him/her in the room is an important step towards a higher-level representation of the meeting dynamics. Similar to gesture recognition, a new distance function is defined and it serves as the core of the algorithm to distinguish between individual speech and overlap speech temporal segments. The overlap speech detection problem is framed as outlier detection problem. An incoming audio is broken into temporal segments based on Bayesian Information Criterion (BIC). Each of these segments is considered as node and conditional distance between the nodes are determined. The underlying distances for triples used in conditional distances is the symmetric KL distance. As each node is modeled as a Gaussian, the distance between the two segments or nodes is given by Monte-Carlo estimation of the KL distance. An MDS based global embedding is created based on the pairwise distance between the nodes and RANSAC is applied to compute the outliers. NIST meeting room data set is used to perform experiments on the overlap speech detection. An improvement of more than 20% is achieved with conditional distance based approach when compared to a KL distance based approach.

A total of 9 brushtail possums (Trichosurus vulpecula) were trained to perform a two-manipulanda, conditional discrimination task. The possums learned to press a right lever in the presence of a tone (80 dB(A)) during tone-on trials, and a left lever in the absence of the tone on tone-off trials. Overall sessions of 11 conditions contained tone frequency between 100 Hz and 35 kHz were tested. Each condition contained training and then probe sessions. In training sessions, the possums were presented with tone-on and tone-off trials, pseudo-randomly. Once the possums responded with over 90% accuracy for five consecutive sessions, then probe sessions were introduced. Probe sessions were similar to the training sessions, except that the tone intensity for tone-on trials was reduced by 8 dB(A) across blocks of 20 trials until their response accuracy in a block fall below 60% or reached 24 dB(A). Data were analysed using overall percentages correct and log d analysis. Both measures indicated that overall response accuracy decreased for all possum as tone intensity reduced. Based on these data analyses, threshold values were calculated using the criterions at 75% correct and a log d of 0.48. The threshold values for each possum and across all possums were plotted as a function of the tone frequency to produce an audiogram. A curvilinear regression was fitted for each threshold values. The functions of both measures were very similar. Both audiograms showed that the possums could hear the tones between 100 Hz and 35 kHz, and were most sensitive to tones between 15 and 20 kHz. This experiment involved many difficulties with producing and measuring tones especially outside of human hearing range. Due to these difficulties, several problems and concerns were raised during the experiment, these were discussed in this study and also recommendations for future research were then presented.

我國廠商自1980年代中期以後，由於台幣升值、工資上漲、環保意識抬頭以及土地取得不易等因素導致競爭力下降，紛紛至國外投資設廠生產，此時台商對外投資的區域大都集中於東南亞和美國。1987年政府開放民眾赴大陸探親，使香港亦成為台商投資大陸的中繼站，1991年政府更直接開放廠商前往大陸直接投資；而日本的投資案件也伴隨市場開放逐漸增加。隨著歐洲聯盟的經濟暨貨幣聯盟Economic and Monetary Union（EMU）的醞釀與開始運作，以及中、東歐的共產封閉經濟體系逐漸向外開放，自1990年起台商也逐漸增加對歐洲的投資。因此，瞭解影響台商對外直接投資區位選擇的因素，是值得研究的課題。 本研究以1988-1997年經濟部投審會所核准的對外投資廠商作為實證研究對象，投資區位則依地主國地緣位置及經濟發展程度劃分為北美洲、歐洲、亞洲已開發國家或新興工業化國家以及大陸與東南亞國家等四大區位。並以McFadden（1974）的Conditional Logit Model作為實證模型，探討勞動成本、資金成本、市場大小、基礎建設、及群聚效果在台商對外直接投資區位的選擇決策中所扮演的角色。 本研究結果發現，在勞動成本方面，低廉的工資對赴大陸與東南亞國家以及歐洲投資的台商有吸引力。在資金成本方面，除了租稅競爭的亞洲國家之外，低稅率是吸引台商前往海外投資的一大因素。在市場要素方面，龐大消費市場的潛在商機不容許台商忽視其存在。基礎建設的良窳僅影響台商對北美洲投資的意願；對其他區域而言，中央政府經濟事務性支出愈高或許代表著當地基礎建設的不足而阻礙了台商的投資意願。至於群聚效果則是影響台商對外投資區位選擇的要素之一，但對於至亞洲已開發或新興工業化國家投資的台商而言，群聚負效果是他們望而卻步的。

關於二維之變數，Arnold and Press (1989) 首先提出檢驗兩個條件分配是否滿足相容性的理論。本研究嘗試對n維之變數，探討n個條件分配滿足相容性的檢驗方式；並提出在三維聯合分配下，給定三個條件分配為常態(normal) 時，檢驗此三個條件分配滿足相容性的充分必要條件；最後，並推導出此三個條件分配滿足相容性時，其所對應的聯合機率密度函數之公式。若此三個條件分配其所對應的聯合機率密度函數進一步假設為常態時，檢驗其相容性的充分必要條件可更加以簡化。 / Arnold and Press (1989) first provide the theory about the compatibility of two conditional distributions in two dimensions. In this research, we extend the two dimensional cases to the high dimensional cases. In particular, we find the necessary and sufficient conditions of the compatibility of three conditional normal distributions in three dimensions. Furthermore, we also provide a formula to find the joint probability density function when three dimensional conditional normal distributions are compatible. Finally, simple sufficient and necessary conditions are also given when the joint distribution is further assumed to be normal.

<p>Using mainly quantile regressions, this paper evaluates the gender wage gap throughout the conditional wage distribution in Sweden. The gender wage is found to increase at the upper tail of the wage distribution, indicating an enforcement of the glass ceiling effect recorded in earlier studies.</p><p>The results also indicate that the earlier noted trend of diminishing wage differences at the bottom of the wage distribution now is turning. The increase of overall wage inequalities coincides with a general increase in wage dispersion among high-income and low-income individuals. It is also noted that there are substantial differences in returns to productivity characteristics between the public and the private sectors, and that both the highest and the lowest unexplained gender wage gap is found in the public sector.</p>

This research develops and applies nonparametric estimation tools in two sectors of interest of financial econometrics: risk management and market microstructure. In the first part we address the problem of estimating conditional quantiles in financial and economic time series. Research in this field received great impulse since quantile based risk measures such as Value at Risk (VaR) have become essential tools to assess the riskiness of trading activities. The great amounts of data available in financial time series allows building nonparametric estimators that are not subject to the risk of specification error of parametric models. A wavelet based estimator is developed. With this approach, minimum regularity conditions of the underlying process are required. Moreover the specific choice of the wavelets in this work leads to the constructions of shape preserving estimators of probability functions. In other words, estimates of probability functions, both densities and cumulative distribution functions, are probability functions themselves. This method is compared with competing methods through simulations and applications to real data. In the second part we carry out a nonparametric analysis of financial durations, that is of the waiting times between particular financial events, such as trades, quote updates, volume accumulation, that happen in financial markets. These data display very peculiar stylized facts one has to take into account when attempting to model them. We make use of an existing algorithm to describe nonparametrically the dynamics of the process in terms of its lagged realizations and of a latent variable, its conditional mean. The estimation devices needed to effectively apply the algorithm to our dataset are presented in this part of the work.