Multilevel analyses in marketing research:differentiating analytical outcomes

Marketing scholars are increasingly recognizing the importance of investigating phenomena at multiple levels. However, the analyses methods that are currently dominant within marketing may not be appropriate to dealing with multilevel or nested data structures. We identify the state of contemporary multilevel marketing research, finding that typical empirical approaches within marketing research may be less effective at explicitly taking account of multilevel data structures than those in other organizational disciplines. A Monte Carlo simulation, based on results from a previously published marketing study, demonstrates that different approaches to analysis of the same data can result in very different results (both in terms of power and effect size). The implication is that marketing scholars should be cautious when analyzing multilevel or other grouped data, and we provide a discussion and introduction to the use of hierarchical linear modeling for this purpose.

The human factors of computer input using a mouse device

This thesis investigates how people select items from a computer display using the mouse input device. The term computer mouse refers to a class of input devices which share certain features, but these may have different characteristics which influence the ways in which people use the device. Although task completion time is one of the most commonly used performance measures for input device evaluation, there is no consensus as to its definition. Furthermore most mouse studies fail to provide adequate assurances regarding its correct measurement.Therefore precise and accurate timing software were developed which permitted the recording of movement data which by means of automated analysis yielded the device movements made. Input system gain, an important task parameter, has been poorly defined and misconceptualized in most previous studies. The issue of gain has been clarified and investigated within this thesis. Movement characteristics varied between users and within users, even for the same task conditions. The variables of target size, movement amplitude, and experience exerted significant effects on performance. Subjects consistently undershot the target area. This may be a consequence of the particular task demands. Although task completion times indicated that mouse performance had stabilized after 132 trials the movement traces, even of very experienced users, indicated that there was still considerable room for improvement in performance, as indicated by the proportion of poorly made movements. The mouse input device was suitable for older novice device users, but they took longer to complete the experimental trials. Given the diversity and inconsistency of device movements, even for the same task conditions, caution is urged when interpreting averaged grouped data. Performance was found to be sensitive to; task conditions, device implementations, and experience in ways which are problematic for the theoretical descriptions of device movement, and limit the generalizability of such findings within this thesis.

Tests of proportional hazards and proportional odds models for grouped survival data

In this paper, we derive score test statistics to discriminate between proportional hazards and proportional odds models for grouped survival data. These models are embedded within a power family transformation in order to obtain the score tests. In simple cases, some small-sample results are obtained for the score statistics using Monte Carlo simulations. Score statistics have distributions well approximated by the chi-squared distribution. Real examples illustrate the proposed tests.

Modeling grouped survival data with time-dependent covariates

In this article, proportional hazards and logistic models for grouped survival data were extended to incorporate time-dependent covariates. The extension was motivated by a forestry experiment designed to compare five different water stresses in Eucalyptus grandis seedlings. The response was the seedling lifetime. The data set was grouped since there were just three occasions in which the seedlings was visited by the researcher. In each of these occasions also the shoot height was measured and therefore it is a time-dependent covariate. Both extended models were used in this example, and the results were very similar.

Seed-based biclustering of gene expression data

Background Accumulated biological research outcomes show that biological functions do not depend on individual genes, but on complex gene networks. Microarray data are widely used to cluster genes according to their expression levels across experimental conditions. However, functionally related genes generally do not show coherent expression across all conditions since any given cellular process is active only under a subset of conditions. Biclustering finds gene clusters that have similar expression levels across a subset of conditions. This paper proposes a seed-based algorithm that identifies coherent genes in an exhaustive, but efficient manner. Methods In order to find the biclusters in a gene expression dataset, we exhaustively select combinations of genes and conditions as seeds to create candidate bicluster tables. The tables have two columns: (a) a gene set, and (b) the conditions on which the gene set have dissimilar expression levels to the seed. First, the genes with less than the maximum number of dissimilar conditions are identified and a table of these genes is created. Second, the rows that have the same dissimilar conditions are grouped together. Third, the table is sorted in ascending order based on the number of dissimilar conditions. Finally, beginning with the first row of the table, a test is run repeatedly to determine whether the cardinality of the gene set in the row is greater than the minimum threshold number of genes in a bicluster. If so, a bicluster is outputted and the corresponding row is removed from the table. Repeating this process, all biclusters in the table are systematically identified until the table becomes empty. Conclusions This paper presents a novel biclustering algorithm for the identification of additive biclusters. Since it involves exhaustively testing combinations of genes and conditions, the additive biclusters can be found more readily.

Fusion of multi resolution remote sensing data for urban sprawl analysis

Urban population is growing at around 2.3 percent per annum in India. This is leading to urbanisation and often fuelling the dispersed development in the outskirts of urban and village centres with impacts such as loss of agricultural land, open space, and ecologically sensitive habitats. This type of upsurge is very much prevalent and persistent in most places, often inferred as sprawl. The direct implication of such urban sprawl is the change in land use and land cover of the region and lack of basic amenities, since planners are unable to visualise this type of growth patterns. This growth is normally left out in all government surveys (even in national population census), as this cannot be grouped under either urban or rural centre. The investigation of patterns of growth is very crucial from regional planning point of view to provide basic amenities in the region. The growth patterns of urban sprawl can be analysed and understood with the availability of temporal multi-sensor, multi-resolution spatial data. In order to optimise these spectral and spatial resolutions, image fusion techniques are required. This aids in integrating a lower spatial resolution multispectral (MSS) image (for example, IKONOS MSS bands of 4m spatial resolution) with a higher spatial resolution panchromatic (PAN) image (IKONOS PAN band of 1m spatial resolution) based on a simple spectral preservation fusion technique - the Smoothing Filter-based Intensity Modulation (SFIM). Spatial details are modulated to a co-registered lower resolution MSS image without altering its spectral properties and contrast by using a ratio between a higher resolution image and its low pass filtered (smoothing filter) image. The visual evaluation and statistical analysis confirms that SFIM is a superior fusion technique for improving spatial detail of MSS images with the preservation of spectral properties.

Ultrasonic Guided Wave Based Damage Detection Using Low-dimensional Data from Laser Doppler Scan

In the current state of the art, it remains an open problem to detect damage with partial ultrasonic scan data and with measurements at coarser spatial scale when the location of damage is not known. In the present paper, a recent development of finite element based model reduction scheme in frequency domain that employs master degrees of freedom covering the surface scan region of interests is reported in context of non-contact ultrasonic guided wave based inspection. The surface scan region of interest is grouped into master and slave degrees of freedom. A finite element wise damage factor is derived which represents damage state over distributed areas or sharp condition of inter-element boundaries (for crack). Laser Doppler Vibrometer (LDV) scan data obtained from plate type structure with inaccessible surface line crack are considered along with the developed reduced order damage model to analyze the extent of scan data dimensional reduction. The proposed technique has useful application in problems where non-contact monitoring of complex structural parts are extremely important and at the same time LDV scan has to be done on accessible surfaces only.

Description of economic data collected with a random sample of commercial reef fish boats in the Florida Keys

This study summarizes the results of a survey designed to provide economic information about the financial status of commercial reef fish boats with homeports in the Florida Keys. A survey questionnaire was administered in the summer and fall of 1994 by interviewers in face-to-face meetings with owners or operators of randomly selected boats. Fishermen were asked for background information about themselves and their boats, their capital investments in boats and equipment, and about their average catches, revenues, and costs per trip for their two most important kinds of fishing trips during 1993 for species in the reef fish fishery. Respondents were characterized with regard to their dependence on the reef fish fishery as a source of household income. Boats were described in terms of their physical and financial characteristics. Different kinds of fishing trips were identified by the species that generated the greatest revenue. Trips were grouped into the following categories: yellowtail snapper (Ocyurus chrysurus); mutton snapper (Lutjanus analis), black grouper (Mycteroperca bonaci), or red grouper (Epinephelus morio); gray snapper (Lutjanus griseus); deeper water groupers and tilefishes; greater amberjack (Seriola dumerili); spiny lobster (Panulirus argus); king mackerel (Scomberomorus cavalla); and dolphin (Coryphaena hippurus). Average catches, revenues, routine trip costs, and net operating revenues per boat per trip and per boat per year were estimated for each category of fishing trips. In addition to its descriptive value, data collected during this study will aid in future examinations of the economic effects of various regulations on commercial reef fish fishermen.(PDF file contains 48 pages.)

Growth of skipjack, Katsuwonus pelamis, and yellowfin, Thunnus albacares, tunas in the Eastern Pacific Ocean, as estimated from tagging data

ENGLISH: The average linear growth rate of skipjack in the eastern Pacific is less than 1 mm per day except for fish 375 to 424 mm in length at release. The growth rate shows a decrease with increasing length and increasing time at liberty. The growth rate of fish in the length range of about 43 to 57 cm is apparently more rapid in the eastern Pacific than in the western Pacific. Dsing data for the northeastern and southeastern Pacific combined, K and ~ were estimated to be 0.658 (on an annual basis) and 885 mm, respectively, by the ungrouped method and 0.829 and 846 mm, respectively, by the grouped method. Sensitivity analyses have shown however, that the estimates of these parameters are poorly determined by the sum of squares method used to derive them. Estimates of K and ~ for the eastern Pacific tend to be lower and higher, respectively, than those for the western Pacific. The average linear growth rate of yellowfin in the eastern Pacific is a little less than 1 mm per day for fish between about 25 and 100 cm in length at release. The growth appears to be most rapid in Area 2 (Revillagigedo Islands) and slowest in Areas 1 (Baja California), 5 (Central America- Colombia), and 6 (Ecuador-Peru). There is considerable variation in the growth rates of individual fish. The growth does not show a decrease with increasing length or increasing time at liberty so realistic estimates of the parameters of the von Bertalanffy or other similar equations cannot be calculated from these data. If realistic estimates of these parameters are to be secured larger fish must be tagged and released or many more long-term returns from fish to about 100 cm in length at release must be obtained. The growth patterns for the eastern Pacific, central Pacific and eastern Atlantic found by most other investigators differ from one another and from those found in the present study. Some of these differences may be real and others may be due to deficiencies in the data or the methods of analysis. Estimates obtained from tagging data are believed to be realistic provided the tags do not inhibit the growth of the fish. It appears that the growth rates of single- and double-tagged fish are the same; this indicates, though not unequivocally, that the tags do not inhibit the growth. SPANISH: La tasa media de crecimiento lineal del barrilete en el Pacífico oriental es inferior a lmm/día, excepto en el caso de peces de entre 375y 424mm de longitud de liberación. La tasa de crecimiento disminuye a medida que aumenta la longitud y el tiempo en libertad. La tasa de crecimiento de peces de entre unos 43 y 57 cm de longitud parece ser mayor en el Pacífico oriental que en el occidental. A partir de datos del Pacífico nororiental y suroriental combinados, se estimaron K y loo en 0.658 (anual) y 885mm, respectivamente, usando el método no agrupado, y 0.829 y 846mm, respectivamente, usando el método agrupado. Sin embargo, los análisis de sensitividad han demostrado que el método de suma de cuadrados utilizado para derivar las estimaciones de estos parámetros las determina con poca precisión. Las estimaciones de K y loo para el Pacífico oriental suelen ser inferiores y superiores, respectivamente, a los del Pacífico occidental. La tasa media de crecimiento lineal del aleta amarilla en el Pacífico oriental es ligeramente inferior a lmm/día para los peces de entre unos 25y 100cmde longitud de liberación. El crecimiento parece ser más rápido en el Area 2(Islas Revillagigedo),y más lento en las Areas 1(Baja California), 5 (Centroamérica-Colombia), y 6 (Ecuador-Perú). Las tasas de crecimiento de peces individuales varían considerablemente. El crecimiento no muestra una disminuciónconun aumento en la longitud o en el tiempo en libertad, y por consecuencia no se se pueden calcular estimaciones realistas de los parámetros de la ecuación de von Bertalanffy u otras ecuaciones similares a partir de estos datos. Para obtener estimaciones realistas de estos parámetros sería necesario marcar peces mayores u obtener muchas más devoluciones a largo plazo de marcas de peces de unos 100cm de longitud de liberación. Los patrones de crecimiento correspondientes al Pacífico oriental, Pacífico central, y Atlántico oriental descubiertos por la mayoría de los investigadores son diferentes entre síy también de los del presente estudio. Es posibleque algunas de estas diferencias sean verdaderas, mientras que otras se deban a faltas en los datos on en los métodos analíticos utilizados. Se considera que las estimaciones obtenidas a partir de los datos de marcado son realistas, suponiendo siempre que las marcas no impidan el crecimiento de los peces. Parece ser que las tasas de crecimiento de peces con una marca y con dos son idénticas, lo cual indica, aunque sin certeza total, que las marcas no ejercen tal efecto. (PDF contains 76 pages.)

CoDaPack. An Excel and Visual Basic based software of compositional data analysis. Current version and discussion form upcoming versions

In the eighties, John Aitchison (1986) developed a new methodological approach for the statistical analysis of compositional data. This new methodology was implemented in Basic routines grouped under the name CODA and later NEWCODA inMatlab (Aitchison, 1997). After that, several other authors have published extensions to this methodology: Marín-Fernández and others (2000), Barceló-Vidal and others (2001), Pawlowsky-Glahn and Egozcue (2001, 2002) and Egozcue and others (2003). (...)

Unsupervised segmentation using Gabor wavelets and statistical features in LIDAR data analysis

In this paper, we address issues in segmentation Of remotely sensed LIDAR (LIght Detection And Ranging) data. The LIDAR data, which were captured by airborne laser scanner, contain 2.5 dimensional (2.5D) terrain surface height information, e.g. houses, vegetation, flat field, river, basin, etc. Our aim in this paper is to segment ground (flat field)from non-ground (houses and high vegetation) in hilly urban areas. By projecting the 2.5D data onto a surface, we obtain a texture map as a grey-level image. Based on the image, Gabor wavelet filters are applied to generate Gabor wavelet features. These features are then grouped into various windows. Among these windows, a combination of their first and second order of statistics is used as a measure to determine the surface properties. The test results have shown that ground areas can successfully be segmented from LIDAR data. Most buildings and high vegetation can be detected. In addition, Gabor wavelet transform can partially remove hill or slope effects in the original data by tuning Gabor parameters.

Karyotypic Data on 28 Species of Scinax (Amphibia: Anura: Hylidae): Diversity and Informative Variation

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Genetic data of 15 autosomal STR loci: an analysis of the Araraquara population colonization (São Paulo, Brazil)

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Parasitological Data as Monitors of Environmental Health

When an appropriate fish host is selected, analysis of its parasites offers a useful, reliable, economical, telescoped indication or monitor of environmental health. The value of that information increases when corroborated by another non-parasitological technique. The analysis of parasites is not necessarily simple because not all hosts serve as good models and because the number of species, presence of specific species, intensity of infections, life histories of species, location of species in hosts, and host response for each parasitic species have to be addressed individually to assure usefulness of the tool. Also, different anthropogenic contaminants act in a distinct manner relative to hosts, parasites, and each other as well as being influenced by natural environmental conditions. Total values for all parasitic species infecting a sample cannot necessarily be grouped together. For example, an abundance of numbers of either species or individuals can indicate either a healthy or an unhealthy environment, depending on the species of parasite. Moreover, depending on the parasitic species, its infection, and the time chosen for collection/examination, the assessment may indicate a chronic or acute state of the environmental health. For most types of analyses, the host should be one that has a restricted home range, can be infected by numerous species of parasites, many of which have a variety of additional hosts in their life cycles, and can be readily sampled. Data on parasitic infections in the western mosquitofish (Gambusia affinis), a fish that meets the criteria in two separate studies, illustrate the usefulness of that host as a model to indicate both healthy and detrimentally influenced environments. In those studies, species richness, intensity of select species, host resistance, other hosts involved in life cycles, and other factors all relate to site and contaminating discharge.