Traditional Experimental Design Research Articles

Perceptual-cognitive expertise in sport: Some considerations when applying the expert performance approach

The number of researchers studying perceptual-cognitive expertise in sport is increasing. The intention in this paper is to review the currently accepted framework for studying expert performance and to consider implications for undertaking research work in the area of perceptual-cognitive expertise in sport. The expert performance approach presents a descriptive and inductive approach for the systematic study of expert performance. The nature of expert performance is initially captured in the laboratory using representative tasks that identify reliably superior performance. Process-tracing measures are employed to determine the mechanisms that mediate expert performance on the task. Finally, the specific types of activities that lead to the acquisition and development of these mediating mechanisms are identified. General principles and mechanisms may be discovered and then validated by more traditional experimental designs. The relevance of this approach to the study of perceptual-cognitive expertise in sport is discussed and suggestions for future work highlighted.

Optimal observation network design for parameter structure identification in groundwater modeling

Designing a pumping observation experiment for parameter identification in groundwater modeling is a very challenging problem because the parameter structure, such as the hydraulic conductivity of a heterogeneous aquifer, is usually unknown. Generally, to identify a more complex parameter structure requires more observation data. Traditional experimental design methods have two disadvantages: First, the sufficiency of information provided by a design is not considered in the design stage. Second, the decision variables often are dependent on the unknown parameter because of the nonlinearity of the model. As a result, a sequential design observation process should be used, but this often is impractical in groundwater modeling. This paper presents a new methodology for experimental design that can find a minimum cost design to provide sufficient information for identifying both the parameter structure and parameter values. Sequential Gaussian simulation is used to generate different realizations for the unknown parameter. For each possible realization the number of observation wells is increased gradually during the design process until the information provided by the design is sufficient. The selection criterion for locating a new observation well is the maximization of the information content for parameter identification. The overall sufficiency of a design is assessed by Monte Carlo simulation. The proposed methodology is explained in detail by a numerical example.

Adaptive combinatorial design to explore large experimental spaces: approach and validation

Systems biology requires mathematical tools not only to analyse large genomic datasets, but also to explore large experimental spaces in a systematic yet economical way. We demonstrate that two-factor combinatorial design (CD), shown to be useful in software testing, can be used to design a small set of experiments that would allow biologists to explore larger experimental spaces. Further, the results of an initial set of experiments can be used to seed further 'Adaptive' CD experimental designs. As a proof of principle, we demonstrate the usefulness of this Adaptive CD approach by analysing data from the effects of six binary inputs on the regulation of genes in the N-assimilation pathway of Arabidopsis. This CD approach identified the more important regulatory signals previously discovered by traditional experiments using far fewer experiments, and also identified examples of input interactions previously unknown. Tests using simulated data show that Adaptive CD suffers from fewer false positives than traditional experimental designs in determining decisive inputs, and succeeds far more often than traditional or random experimental designs in determining when genes are regulated by input interactions. We conclude that Adaptive CD offers an economical framework for discovering dominant inputs and interactions that affect different aspects of genomic outputs and organismal responses.

Tool life testing in gundrilling: an application of the group method of data handling (GMDH)

Although the known techniques of the design of experiments (DOE) allows significant improvements in the methodology of machining tests, their use in metal cutting is restricted by limited number of variables included into consideration, pre-set model, subjective pre-process decisions and uncertainties in the machining data collected experimentally. This paper introduces an application of a new powerful technique of DOE known as the group method of data handling (GMDH) to the design of tool life experiments that allows avoiding the above-mentioned disadvantages of traditional DOE. As a result of such an application, a mathematical model that correlates tool life in gundrilling of cast iron with various regime and design parameters was obtained. Special care was taken to avoid the influence of the dynamic phenomenon of the gundrilling process on the obtained experimental results. The obtained mathematical model reveals that tool life in gundrilling is a complex function of various regime and design parameters.

Evaluation of the Taguchi methods for the simultaneous assessment of the effects of multiple variables in the tumour microenvironment

BackgroundThe control of proliferation, differentiation and survival of normal and malignant cells in the tumour microenvironment is under the control of a wide range of different factors, including cell:cell interactions, cytokines, growth factors and hormonal influences. However, the ways in which these factors interact are poorly understood. In order to compare the effects of multiple variables, experimental design becomes complex and difficult to manage. We have therefore evaluated the use of a novel approach to multifactorial experimental design, the Taguchi methods, to approach this problem.MethodThe Taguchi methods are widely used by quality engineering scientists to compare the effects of multiple variables, together with their interactions, with a simple and manageable experimental design. In order to evaluate these methods, we have used a simple and robust system to compare a traditional experimental design with the Taguchi Methods. The effect of G-CSF, GM-CSF, IL3 and M-CSF on daunorubicin mediated cytotoxicity in K562 cells was measured using the MTT assay.ResultsBoth methods demonstrated that the same combination of growth factors at the same concentrations minimised daunorubicin cytotoxicity in this assay.ConclusionsThese findings demonstrate that Taguchi methods may be a valuable tool for the investigation of the interactions of multiple variables in the tumour microenvironment.

Dolphin echolocation strategies studied with the Biosonar Measurement Tool

Two free-swimming dolphins (Tt722 and Tt673) were trained to carry the Biosonar Measurement Tool (BMT) during open water, proud target searches in order to explore echolocation behavior without the constraints of traditional experimental designs. The BMT recorded the angular motion, depth, and velocity of the dolphin as well as echolocation clicks and echoes returning from insonified targets. Mean search time for Tt722 was 24.6±7.3 s and 6.5±3.0 s for Tt673 on target present trials, the former strategy resulting in the lower false alarm rate. The majority of clicks exceeded 195 dB re: 1 μPa throughout all trials for both animals but each demonstrated preferences for particular frequency bands of echolocation. Considering all trials, only 3.6% of all clicks produced by Tt722 contained peak frequencies greater than 60 kHz whereas Tt673 produced clicks with peak frequencies above 60 kHz 20.4% of the time. Distinctive frequency bands in the distribution of clicks were notable: bands for Tt673 occurred at 38, 54, and 69 kHz with less defined higher order bands; bands for Tt722 occurred at 25, 35, and 40 kHz. Distinctive frequency bands suggest a preferential use or mechanical constraint on harmonically related click frequencies.

General balance, large data sets and extensions to unbalanced treatment structures

General balance defines an important class of designs covering virtually all the traditional experimental designs and, in particular, those with several error terms. In these designs the block terms are mutually orthogonal, the treatment terms are also mutually orthogonal, and the contrasts of each treatment term have equal efficiency factors in each of the strata where they are estimated. Generally balanced designs can efficiently be analysed by an algorithm which involves a series of sweep operations for each stratum. Each sweep estimates the effects of a treatment term (by its means divided by the efficiency factor), and subtracts them from the current working vector which then becomes the working vector for the next sweep. Workspace requirements are considerably less than those of multiple-regression style algorithms, so the algorithm has great advantages particularly with large treatment models. With large data sets, however, it may be more difficult to fulfill all the conditions of general balance. One possibility is to retain orthogonal block structure, but allow unbalanced treatment structures. It is shown that the algorithm can successfully be adapted to this situation, but it now requires an iterative sequence of sweeps to fit the treatment model in each stratum.

Sequential robust design methodology and X-ray photoelectron spectroscopy to analyze the grafting of hyaluronic acid to glass substrates.

Sequential Robust Design experiments and X-ray photoelectron spectroscopic (XPS) studies were performed to examine the immobilization of hyaluronic acid (HA) on glass substrates chemisorbed with N-(2-aminoethyl)-3-aminopropyl-trimethoxysilane (EDS). Numerous reaction conditions were investigated, including the concentrations of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysulfosuccinimide (Sulfo-NHS), and HA, and the reaction buffer type, concentration, and pH. The elemental surface compositions of carbon and silicon (C/Si ratio) were used to assess the extent of HA immobilization, leading to the identification of critical HA-binding reaction conditions and the determination of an optimum surface chemistry. The optimum chemistry consisted of 200 mM EDC, 50 mM Sulfo-NHS, 10 mM N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) (HEPES) buffer at a pH of 7.0, and 3 mg/mL HA. This work emphasizes the advantages of using Robust Design methods over traditional statistical experimental design, particularly when large numbers of variables are examined and costly analytical techniques are employed.

Combinatorial methodologies offer potential for rapid research of photoresist materials and formulations

A combinatorial research methodology is outlined with the potential to investigate material factors in photoresist formulations more rapidly than with traditional experimental design. The approach involves generating a gradient of processing variables in the photoresist film and using a set of analysis tools to “map” the photoresist properties as a function of the gradient variables. While high-throughput strategies have proven useful for empirical optimization of photoresist processing conditions, they have not been fully exploited to probe the complex array of material issues in projection lithography. Of primary importance in combinatorial methodologies is the generation of variable gradients. The focus of this article is to illustrate techniques to generate a gradient in bake temperature and bake time in the polymer films, and to illustrate that these gradient techniques can be used to probe fundamental properties of lithographic materials, such as deprotection kinetics. These gradient techniques were applied to rapidly probe the deprotection behavior of a poly(tertbutyloxy-carbonyloxy-styrene). The deprotection data, obtained by applying the combinatorial gradients, agrees well with previous literature, validating the accuracy of these high throughput methods. The advantage of the combinatorial research strategies lies in the potential to investigate fundamental lithographic phenomena more rapidly than with traditional experimentation.

Approaches to understanding susceptibility to nephropathy: From genetics to genomics

The incidence of end-stage renal disease (ESRD) is increasing worldwide despite efforts to slow the progression of chronic renal failure (CRF) by controlling blood pressure and hyperglycemia. Two available therapies for ESRD, dialysis and transplantation, are expensive and are at best palliative. Recently, data from several laboratories have demonstrated that ESRD is under substantial genetic control, and efforts to identify these genetic determinants are underway. Identifying genes for ESRD pathogenesis has several goals. First, understanding the genetic basis of ESRD offers a means to clarify the mechanisms that result in kidney pathobiology. Second, better and new treatments for prevention of progression of CRF to ESRD may be developed. Third, individuals at risk could be identified early in their course and targeted for intensive therapy. Finally, the products of genes causing disease become target molecules for gene therapy. In this article, we discuss data from our laboratories, which employ two different molecular genetic strategies for identifying ESRD pathogenesis genes. In contrast to traditional experimental design, both approaches are hypothesis generating, identifying candidate molecules for further study, rather than hypothesis driven and may provide novel insights into mechanisms of renal disease progression.

Exploration and Selection of Intralimb Coordination Patterns in 3-Month-Old Infants

Through the exploration of their own capacities and the selection of adaptive responses, infants learn new motor solutions. Using a conjugate reinforcement mobile procedure, previous researchers have repeatedly shown that infants increase their leg kick frequency to control a mobile that is connected to their ankles. That traditional experimental design allows multiple motor solutions to the task and therefore provides limited information about the infants' capacity to explore and select specific motor solutions. The author designed a new experimental procedure to study infants' capacity to discover and adopt specific motor solutions. The new, constraining mobile reinforcement procedure requires a specific motor response and therefore the development of a more finely tuned perception—action map than has previously been experimentally demonstrated. To gain reinforcement from the mobile, infants had to produce a coordinated hip and knee extension within the same leg. The results from the 13 infant participants showed that they were capable of increasing their frequency of coordinated movements to make the mobile move. Those results suggest that infants at the age of 89–106 days are sensitive to intralimb coordination task requirements and are capable of mapping their own limb dynamics to the environmental information.

Evaluation of Teaching and Learning Strategies

With the growing awareness of the importance of teaching and learning in universities and the need to move towards evidence-based teaching, it behooves the professions to re-examine their educational research methodology. While the what, how and why of student learning have become more explicit, the professions still struggle to find valid methods of evaluating the explosion of new innovation in teaching/learning strategies. This paper discusses the problems inherent in applying traditional experimental design techniques to advances in educational practice.

Response Surface Experimental Designs for Investigating Interspecific Competition

Competition is an important interaction in ecology, and many experiments have been done to examine the effects of intraspecific and interspecific competition. Unfortunately, most of these experiments have been done using either substitution or additive experimental designs. Substitution designs, and to a lesser extent additive designs, severely limit the useful inferences that ecologists can draw from the resulting data. Response surface experimental designs, which vary the densities of two competing species independently, offer the advantage of being able to compare the fits of alternative competition models and to estimate model parameters. This kind of experimental design has been relatively little used in ecology to date. I consider the use of response surface designs from two perspectives. The first is the optimal allocation of experimental effort among density treatments, with the goal of distinguishing among alternative models. Second, I compare the ability of six response surface designs to estimate known competition coefficients correctly, using simulated data. Low densities and densities near equilibrium in one generation tend to return similar densities the following generation and thus provide little or no information for comparing alternative population models. Experimental designs that span a wide range of densities, including high densities above carrying capacities, provide more accurate parameter estimates in general, but low densities are helpful for estimating population growth rates. Designs that are not fully factorial are one compromise solution that can yield accurate parameter estimates for models and yet require less experimental effort than full factorial surfaces. In general, response surfaces are a powerful class of experimental design and offer potentially stronger connections between empirical and theoretical approaches than traditional experimental designs.

RESPONSE SURFACE EXPERIMENTAL DESIGNS FOR INVESTIGATING INTERSPECIFIC COMPETITION

Competition is an important interaction in ecology, and many experiments have been done to examine the effects of intraspecific and interspecific competition. Un- fortunately, most of these experiments have been done using either substitution or additive experimental designs. Substitution designs, and to a lesser extent additive designs, severely limit the useful inferences that ecologists can draw from the resulting data. Response surface experimental designs, which vary the densities of two competing species independently, offer the advantage of being able to compare the fits of alternative competition models and to estimate model parameters. This kind of experimental design has been relatively little used in ecology to date. I consider the use of response surface designs from two perspectives. The first is the optimal allocation of experimental effort among density treatments, with the goal of distinguishing among alternative models. Second, I compare the ability of six response surface designs to estimate known competition coefficients correctly, using sim- ulated data. Low densities and densities near equilibrium in one generation tend to return similar densities the following generation and thus provide little or no information for comparing alternative population models. Experimental designs that span a wide range of densities, including high densities above carrying capacities, provide more accurate pa- rameter estimates in general, but low densities are helpful for estimating population growth rates. Designs that are not fully factorial are one compromise solution that can yield accurate parameter estimates for models and yet require less experimental effort than full factorial surfaces. In general, response surfaces are a powerful class of experimental design and offer potentially stronger connections between empirical and theoretical approaches than traditional experimental designs.

Interspecific Competition in Plants: How Well Do Current Methods Answer Fundamental Questions?

Accurately quantifying and interpreting the processes and outcomes of competition among plants is essential for evaluating theories of plant community organization and evolution. We argue that many current experimental approaches to quantifying competitive interactions introduce size bias, which may significantly impact the quantitative and qualitative conclusions drawn from studies. Size bias generally arises when estimates of competitive ability are erroneously influenced by the initial size of competing individuals. We employ a series of quantitative thought experiments to demonstrate the potential for size bias in analysis of four traditional experimental designs (pairwise, replacement series, additive series, and response surfaces) either when only final measurements are available or when both initial and final measurements are collected. We distinguish three questions relevant to describing competitive interactions: Which species dominates? Which species gains? and How do species affect each other? The choice of experimental design and measurements greatly influences the scope of inference permitted. Conditions under which the latter two questions can give biased information are tabulated. We outline a new approach to characterizing competition that avoids size bias and that improves the concordance between research question and experimental design. The implications of the choice of size metrics used to quantify both the initial state and the responses of elements in interspecific mixtures are discussed. The relevance of size bias in competition studies with organisms other than plants is also discussed.

Interspecific Competition in Plants: How Well Do Current Methods Answer Fundamental Questions?

Accurately quantifying and interpreting the processes and outcomes of competition among plants is essential for evaluating theories of plant community organization and evolution. We argue that many current experimental approaches to quantifying competitive interactions introduce size bias, which may significantly impact the quantitative and qualitative conclusions drawn from studies. Size bias generally arises when estimates of competitive ability are erroneously influenced by the initial size of competing individuals. We employ a series of quantitative thought experiments to demonstrate the potential for size bias in analysis of four traditional experimental designs (pairwise, replacement series, additive series, and response surfaces) either when only final measurements are available or when both initial and final measurements are collected. We distinguish three questions relevant to describing competitive interactions: Which species dominates? Which species gains? and How do species affect each other? The choice of experimental design and measurements greatly influences the scope of inference permitted. Conditions under which the latter two questions can give biased information are tabulated. We outline a new approach to characterizing competition that avoids size bias and that improves the concordance between research question and experimental design. The implications of the choice of size metrics used to quantify both the initial state and the responses of elements in interspecific mixtures are discussed. The relevance of size bias in competition studies with organisms other than plants is also discussed.

Using Collaborative Teacher Research to Determine the Impact of Professional Development School Activities on Elementary Students' Math and Writing Outcomes

As a result of the emphasis on professional development schools (PDSs) in teacher education over the past decade, numerous theoretical articles, process descriptions, and research reports provide considerable information about the nature and impact of school-university partnerships (Abdal-Haaq, 1998; Book, 1996). Several articles outline the perceived benefits for school and university faculty of participation in joint inquiry (e.g., Darling-Hammond, 1992). The literature provides particular insights into what happens to teachers, future teachers, and to a lesser extent, university professors; and how schools and universities change as a result of their collaboration. Most studies focus on roles and relationships, teacher attitudes, and teacher education, rather than on the elementary and secondary students in PDSs (Book, 1996). Although almost all PDS partnership goals include a stated emphasis on improved school achievement and student learning (Freeman, 1996), few studies address the impact of PDS efforts on student outcomes (Abdal-Haqq, 1998; Book, 1996; Valli, Cooper, & Frankes, 1997). Why have researchers not adequately addressed the impact of PDS on K-12 student outcomes? Field research that assesses student learning is difficult and fraught with pitfalls that threaten its rigor. Easy-to-obtain measures such as standardized tests may be too far removed from the focus and activities of PDSs to provide adequate measurement of impact on student learning. Reform efforts focusing on restructuring typically seek to evaluate authentic student activities that have relevance and meaning for them (Newmann, 1991) but may be difficult to measure. Furthermore, it is difficult to isolate distinct variables in complex interventions such as PDSs that can be directly related to student outcomes. Use of control group or other traditional experimental designs, which might serve to identify cause-and-effect relationships, may be seen as unethical by teachers and administrators because they deny potentially beneficial treatments from groups of students and teachers in the school. Another possible reason for the paucity of research on the effects of PDSs on K-12 students involves issues of trust and collaboration between practitioners and researchers. Many opportunities for inquiry to be misunderstood exist in partnerships where tenuous relationships are evolving between two very different cultures struggling with issues of trust, power, and control (Knight, Wiseman, & Smith, 1992; Ruscoe, Whitford, Egginton, & Esselman, 1989). Teachers may have concerns about the implications of experimentation in their classes because the experience may be damaging to students (Darling Hammond, 1992). Furthermore, partnership research can directly affect the professional lives of teachers and administrators. For example, teachers may see research as competing with instructional goals (Book, 1996); less than outstanding findings may reflect badly on programs, teachers, and administrators increasingly under public scrutiny. Teachers and administrators have traditionally been the objects of study by university researchers who may or may not even share their findings with the objects of their research. Teachers often see little relevance in studies that answer questions they do not ask and reports that use terminology they do not understand. As a result, the gap between research and practice widens. One response to the problems of mediating research and practice and excluding teachers' voices in the development of knowledge about teaching and learning is the collaboration of school and university researchers in determining PDS impact. Collaborative inquiry (Sirotnik 1988), also known as teacher research (Hollingsworth & Sockett, 1994), action research (Oja & Pine, 1987; Reason 1994), or collaborative research (Lee, 1993), holds significant promise. The goal of collaborative teacher research is to link research with practice to affect teacher thinking and instructional behavior, school systems and culture, and student outcomes. …

Attribute-Treatment Interaction Research Using Nonexperimental Data and the Allocation Average

Attribute-treatment interaction (ATI) research attempts to answer the question, “Which treatment is best for whom?” Recent critical reviewers of ATI methodology have cited numerous pragmatic and methodological shortcomings of traditional experimental designs and argued that successful ATI research will require an alternative approach. This manuscript describes a purely correlational ATI method that addresses the more serious criticisms of experimental ATI research. This paper also presents use of the allocation average for ATI research that is either experimental or correlational in nature. The allocation average estimates the clinical utility of differential treatment assignment based on ATI data in terms of improved treatment outcomes. The proposed method is illustrated with hypothetical psychotherapy data and data from a recent educational study.

Design keys, pseudo-factors and general balance

Design keys provide an effective way of defining various types of generallybalanced design. This is an important class which coversvirtually all the traditional experimental designs and, inparticular, those with more than one error term (or stratum). They are efficient to analyse and the results arestraightforward to interpret. We describe algorithms that can beused to determine the structure of a design (that is the stratawhere each set of treatment contrasts is estimated and thecorresponding efficiency factors) from the design keys used togenerate the design, and to define the pseudo-factors that maybe necessary to produce the analysis.

Orthogonal Column Latin Hypercubes and Their Application in Computer Experiments

Abstract Latin hypercubes have been frequently used in conducting computer experiments. In this paper, a class of orthogonal Latin hypercubes that preserves orthogonality among columns is proposed. Applying an orthogonal Latin hypercube design to a computer experiment benefits the data analysis in two ways. First, it retains the orthogonality of traditional experimental designs. The estimates of linear effects of all factors are uncorrelated not only with each other, but also with the estimates of all quadratic effects and bilinear interactions. Second, it can facilitate nonparametric fitting procedures, because one can select good space-filling designs within the class of orthogonal Latin hypercubes according to selection criteria.