3 resultados para multivariate data
em Digital Commons at Florida International University
Resumo:
Dissolved organic matter (DOM) in groundwater and surface water samples from the Florida coastal Everglades were studied using excitation–emission matrix fluorescence modeled through parallel factor analysis (EEM-PARAFAC). DOM in both surface and groundwater from the eastern Everglades S332 basin reflected a terrestrial-derived fingerprint through dominantly higher abundances of humic-like PARAFAC components. In contrast, surface water DOM from northeastern Florida Bay featured a microbial-derived DOM signature based on the higher abundance of microbial humic-like and protein-like components consistent with its marine source. Surprisingly, groundwater DOM from northeastern Florida Bay reflected a terrestrial-derived source except for samples from central Florida Bay well, which mirrored a combination of terrestrial and marine end-member origin. Furthermore, surface water and groundwater displayed effects of different degradation pathways such as photodegradation and biodegradation as exemplified by two PARAFAC components seemingly indicative of such degradation processes. Finally, Principal Component Analysis of the EEM-PARAFAC data was able to distinguish and classify most of the samples according to DOM origins and degradation processes experienced, except for a small overlap of S332 surface water and groundwater, implying rather active surface-to-ground water interaction in some sites particularly during the rainy season. This study highlights that EEM-PARAFAC could be used successfully to trace and differentiate DOM from diverse sources across both horizontal and vertical flow profiles, and as such could be a convenient and useful tool for the better understanding of hydrological interactions and carbon biogeochemical cycling.
Resumo:
This study investigated the feasibility of using qualitative methods to provide empirical documentation of the long-term qualitative change in the life course trajectories of “at risk” youth in a school based positive youth development program (the Changing Lives Program—CLP). This work draws from life course theory for a developmental framework and from recent advances in the use of qualitative methods in general and a grounded theory approach in particular. Grounded theory provided a methodological framework for conceptualizing the use of qualitative methods for assessing qualitative life change. The study investigated the feasibility of using the Possible Selves Questionnaire-Qualitative Extension (PSQ-QE) for evaluating the impact of the program on qualitative change in participants' life trajectory relative to a non-intervention control group. Integrated Qualitative/Quantitative Data Analytic Strategies (IQ-DAS) that we have been developing a part of our program of research provided the data analytic framework for the study. ^ Change was evaluated in 85 at risk high school students in CLP high school counseling groups over three assessment periods (pre, post, and follow-up), and a non-intervention control group of 23 students over two assessment periods (pre and post). Intervention gains and maintenance and the extent to which these patterns of change were moderated by gender and ethnicity were evaluated using a mixed design Repeated Measures Multivariate Analysis of Variance (RMANOVA) in which Time (pre, post) was the within (repeated) factor and Condition, Gender, and Ethnicity the between group factors. The trends for the direction of qualitative change were positive from pre to post and maintained at the year-end follow-up. More important, the 3-way interaction for Time x Gender x Ethnicity was significant, Roy's Θ =. 205, F(2, 37) = 3.80, p <.032, indicating that the overall pattern of positive change was significantly moderated by gender and ethnicity. Thus, the findings also provided preliminary evidence for a positive impact of the youth development program on long-term change in life course trajectory, and were suggestive with respect to the issue of amenability to treatment, i.e., the identification of subgroups of individuals in a target population who are likely to be the most amenable or responsive to a treatment. ^
Resumo:
Multivariate normal distribution is commonly encountered in any field, a frequent issue is the missing values in practice. The purpose of this research was to estimate the parameters in three-dimensional covariance permutation-symmetric normal distribution with complete data and all possible patterns of incomplete data. In this study, MLE with missing data were derived, and the properties of the MLE as well as the sampling distributions were obtained. A Monte Carlo simulation study was used to evaluate the performance of the considered estimators for both cases when ρ was known and unknown. All results indicated that, compared to estimators in the case of omitting observations with missing data, the estimators derived in this article led to better performance. Furthermore, when ρ was unknown, using the estimate of ρ would lead to the same conclusion.