Effect of violation of normality on the t test of the correlation coefficient.

Performed simulations to study the effects of violations of the normality assumption on the t test of the Pearson product-moment correlation coefficient. The test was robust as long as the variables were independent, even with violations in combined distributions. Large effects for violations of normality were found when the variables were not independent, even though the population correlation was zero. (18 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Sampling variance in attenuated correlation coefficients: A Monte Carlo study.

Research in validity generalization has generated renewed interest in the sampling error of the Pearson correlation coefficient. The standard estimator for the sampling variance of the correlation was derived under assumptions that do not consider the presence of measurement error or range restriction in the data. The accuracy of the estimator in attenuated or restricted data has not been studied. This article presented the results of computer simulations that examined the accuracy of the sampling variance estimator in data containing measurement error. Sample sizes of n?=?25, n?=?60, and n?=?100 are used, with the reliability ranging from .10 to 1.00, and the population correlation ranging from .10 to 0.90. Results demonstrated that the estimator has a slight negative bias, but may be sufficiently accurate for practical applications if the sample size is at least 60. In samples of this size, the presence of measurement error does not add greatly to the inaccuracy of the estimator. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Contrast effects in paired comparisons: Evidence for both stimulus-based and response-based processes.

Participants made paired comparisons of square sizes, with blue squares presented on the left side of a computer screen and red squares on the right. Context was manipulated by varying the distributions of blue and red squares separately. In Exp 1, target squares from low-range distributions were judged larger than the same-size squares from high-range distributions. In Exp 2, ranges were equated but ranks manipulated between distributions, pairing bell with U-shaped and positively skewed with negatively skewed distributions. Results provided little support for a rank-dependent valuation model. Exp 3 used distributions that were designed to test between adaptation-level and response-equalization models. Both models received support, with response latencies constituting an important moderating variable. Response patterns for short latency participants were consistent with a stimulus-based adaptation-level process, and those for longer latency participants were consistent with a response-equalization process. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Investigation of the robust rank-order test for non-normal populations with unequal variances: The case of reaction time.

Evaluated the performance of the robust rank-order (Fligner-Policello) test of treatment effects for populations with unequal variances. Both symmetric (normal) and skewed (ex-Gaussian) distributions were examined. The ex-Gaussian distribution is particularly relevant because it is representative of skewed distributions found in experimental psychology, most notably reaction time data. Results indicate that when testing the hypothesis of equal medians for the symmetric distribution, the Fligner-Policello test is conservative, whereas it performs inconsistently for the skewed population. In light of the findings, the authors discuss the various options presently available to researchers confronted with unequal variances and skewed data, including a discussion of whether testing for treatment effects for populations with unequal variances is of practical relevance. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Multiple comparison procedures for trimmed means.

Stepwise multiple comparison procedures (MCPs) based on least squares and trimmed estimators were compared for their rates of Type I error and their ability to detect true pairwise group differences. The MCPs were compared in unbalanced one-way completely randomized designs when normality and homogeneity of variance assumptions were violated. Results indicated that MCPs based on trimmed means and Winsorized variances controlled rates of Type I error, whereas MCPs based on least squares estimators typically could not, particularly when the data were highly skewed. However, MCPs based on least squares estimators were substantially more powerful than their counterparts based on trimmed means and Winsorized variances when the data were only moderately skewed, a finding which qualifies recommendations on the use of trimmed estimators offered in the literature. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Modeling microbial populations with the original and modified versions of the continuous and discrete logistic equations

The life histories of microbial populations, under favorable and adverse conditions, exhibit a variety of growth, decay, and fluctuation patterns. They have been described by numerous mathematical models that varies considerably in structure and number of constants. The continuous logistic equation alone and combined with itself or with its mirror image, the Fermi function, can produce many of the observed growth patterns. They include those that are traditionally described by the Gompertz equation and peaked curves, with the peak being symmetric or asymmetric narrow or wide. The shape of survival and dose response curves appears to be determined by the distribution of the resistance's to the lethal agent among the individual organisms. Thus, exponential decay and Fermian or Gompertz-type curves can be considered manifestations of skewed to the right, symmetric, and skewed to the left distributions, respectively. Because of the mathematical constraints and determinism, the original discrete logistic equation can rarely be an adequate model of real microbial populations. However, by making its proportionality constant a normal-random variate it can simulate realistic histories of fluctuating microbial populations, including scenarios of aperiodic population explosions of varying intensities of the kind found in food-poisoning episodes.     

Inter-observer reliability of the Spondylitis Functional Index Instrument for assessing spondylarthropathies

OBJECTIVE: To study the Spondylitis Functional Index (SFI) by having two physical therapists observe patients with spondylitis perform various tasks listed on the instrument. The physical therapists' observations were compared with each other and with the self-reported abilities of the patients. METHODS: Subjects (n = 30) were recruited from a cross-section of patients participating in a prospective randomized, multicenter, double-blind, parallel clinical trial of the efficacy of sulfasalazine on ankylosing spondylitis (n = 13), psoriatic arthritis (n = 13), and Reiter's syndrome (n = 4) conducted at the Veterans Affairs Medical Center in Salt Lake City. Percents of agreement and Cohen's kappa analysis were used to assess the reliability of the observations of the therapists and patients. RESULTS: The overall percent of agreement between the observers on the SFI was 93%. The overall percent of agreement between observer 1 and patients on the SFI was 66% and between observer 2 and patients was 67%. The overall inter-observer reliability measured by the Pearson coefficient was 0.91 and by Cohen's kappa was 0.86. Between observer 1 and the patients the Pearson was r = 0.53 and kappa = 0.39. For observer 2 the Pearson was r = 0.52 and kappa 0.39. CONCLUSIONS: We consider the agreement and reliability between observers to be high. The agreement and inter-observer reliability was poor between observers and patients. The SFI, as enhanced for use in this study to assess change in functional ability of patients with spondylitis, demonstrated high reliability when used by trained observers.     

Resampling techniques in the analysis of non-binormal ROC data

The methods most commonly used for analyzing receiver operating characteristic (ROC) data incorporate "binormal" assumptions about the latent frequency distributions of test results. Although these assumptions have proved robust to a wide variety of actual frequency distributions, some data sets do not "fit" the binormal model. In such cases, resampling techniques such as the jackknife and the bootstrap provide versatile, distribution-independent, and more appropriate methods for hypothesis testing. This article describes the application of resampling techniques to ROC data for which the binormal assumptions are not appropriate, and suggests that the bootstrap may be especially helpful in determining confidence intervals from small data samples. The widespread availability of ever-faster computers has made resampling methods increasingly accessible and convenient tools for data analysis.     

Latent structure of the alexithymia construct: A taxometric investigation.

Despite a wealth of research on the validity of alexithymia and its association with a number of common medical and psychiatric disorders, the fundamental question of whether alexithymia is best conceptualized as a dimensional or categorical construct remains unresolved. In the current investigation, taxometric analysis is used to examine the nature of the latent structure of alexithymia. Several nonredundant taxometric procedures were performed with item sets from the 20-item Toronto Alexithymia Scale (R. M. Bagby, J. D. A. Parker, & G. J. Taylor, 1994) as indicators. These procedures were applied separately in large community (n = 1,933) and undergraduate (n = 1,948) samples and in a smaller sample of psychiatric outpatients (n = 302). The results across various taxometric procedures and the different samples provide strong support that alexithymia is a dimensional construct. Some theoretical implications of these findings for research on the alexithymia construct are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Sensitivity and specificity of antigen-capture ELISAs for diagnosis of Trypanosoma congolense and Trypanosoma vivax infections in cattle

Sensitivity and specificity of the FAO/IAEA antigen-ELISA kits for diagnosis of bovine trypanosomosis were investigated using sera from experimental cattle infected by tsetse challenge with cloned populations of Trypanosoma congolense (three populations) or T. vivax (one population). The kits are based on monoclonal antibodies that recognise internal antigens of tsetse-transmitted trypanosomes. Ten cattle were infected with each trypanosome population for at least 60 days, and in combination with uninfected cohorts (n = 16) were used in a double-blind study design. Sensitivity and specificity of the tests depended on the choice of positive-negative thresholds expressed as percent positivity with respect to the median OD of four replicates of the strong positive reference serum provided with the kit. In general, while overall specificities were high, sensitivities of the antigen-ELISAs were poor. For example, at a cut-off of 5% positivity, the sensitivities of the antigen-ELISAs were 11% for samples (n = 1162) from T. congolense infected cattle (n = 30), and 24% for samples (n = 283) from T. vivax infected cattle (n = 10). The corresponding specificity values were 95% and 79%, respectively. At a cut-off of 2.5% positivity sensitivity for T. congolense was 25%, and for T. vivax 35%; corresponding specificity values were 85% and 63% respectively. There were no values of the positive-negative threshold at which both sensitivity and specificity were satisfactory. Restricting the analyses to samples taken more than 2 weeks after tsetse challenge did little to improve sensitivity estimates. Trypanosome species specificities of the antigen-ELISAs were also poor. Sensitivity and species specificity of the antigen-ELISA for Trypanosoma brucei infections were not investigated. In contrast to the antigen-ELISA, the sensitivity of the buffy-coat technique when applied to the same experimental animals was fairly high at 67% for T. congolense infections and 60% for T. vivax infections. For samples taken more than 2 weeks after tsetse challenge, high sensitivity estimates of 96% for T. congolense and 76% for T. vivax infections were obtained.     

The importance of human glandular kallikrein and its correlation with different prostate specific antigen serum forms in the detection of prostate carcinoma

BACKGROUND: Human glandular kallikrein (hK2), the prostate specific antigen (PSA) close homologue, possesses approximately 80% structure identity with PSA. The identification of PSA was an important step in the detection of prostate carcinoma (PCa). Thus, hK2 measurement in the serum has the potential to become another important diagnostic test for PCa. In the current study, the authors measured the serum concentrations of the hK2 with "in-house" immunofluorometric assays in different patient groups. The correlation between serum hK2 and different PSA forms was investigated. METHODS: The prospectively collected serum samples were obtained preoperatively on admission from 311 consecutive male patients. Sixteen patients did not fulfill inclusion criteria; the remaining patients were divided into four groups (Groups I-III confirmed histologically): Group I: patients with PCa (n = 56); Group II: patients with benign prostatic hyperplasia (BPH) (n = 163); Group III: patients with BPH with a chronic in-dwelling catheter (BPH cat) (n = 44); and Group IV-control group (n = 32). The patients in Group IV had urolithiasis, varicocele, or kidney or bladder tumors). An experimental immunofluorometric assay with an analytic sensitivity of 0.01 ng/mL and a functional sensitivity of 0.05 ng/mL was used to determine serum hK2 concentrations. Total PSA, free PSA, and PSA complexed to alpha-1-antichymotrypsin (PSA-ACT) also were measured. hK2 concentrations equal to or above the functional sensitivity limit were correlated with each of these PSA serum forms. Free to total PSA, hK2 to total PSA, and hK2 to free PSA ratios were calculated and compared in different patient groups. RESULTS: The hK2 concentrations were equal to or above the functional sensitivity limit in 179 of 311 samples (57.6%). In these samples, hK2 correlated best with free PSA (correlation coefficient [r] = 0.79) and correlated well with total PSA (r = 0.72) and PSA-ACT (r = 0.74). Similar correlations also could be observed when each clinical group was analyzed separately. The median proportion of hK2 in relation to total PSA was 2.1%, 1.8%, and 1.4%, respectively, for PCa, BPH, and BPH cat patients. Both the free to total PSA ratio and the hK2 to free PSA ratio discriminated well between PCa and BPH patients. Within the range of total PSA of 4-10 ng/mL (PCa [n = 11] and BPH [n = 41]) the hK2 to free PSA ratio had a specificity of 63.4% and 90.9% sensitivity (area under the receiver operating characteristic [ROC] curve = 0.85) whereas the free to total PSA ratio had a 34.1% specificity at the same sensitivity level (area under ROC curve = 0.74). CONCLUSIONS: The hK2 serum level correlates well with all PSA serum forms in all clearly defined clinical groups. The preliminary finding that the hK2 to free PSA ratio appeared to improve the detection of PCa compared with the free to total PSA ratio in patients with total PSA within a 4-10 ng/mL range is of clinical interest. Combining human serine proteases in the multivariate regression analysis will be a tool to improve cancer detection. Further investigations with more sensitive hK2 assays and in larger patient populations are needed to confirm this finding.     

The variability of mitochondrial DNA in eastern Slavs

Geographical distributions of three mitochondrial DNA markers (BamHI-1/AvaII-5(3), BamHI-2/AvaII-5(3), and BamHI-3/MspI-4) were analyzed in Caucasian populations of Europe and Middle East. Considerable genetic similarity between eastern Slavs and southern European populations (r = 0.995) has been found. It is assumed that the basic components of the mitochondrial gene pool of eastern Slavs are of southern European origin.     

Relation between the osmolality trend and ornithynedecarboxylase activity in red blood cells of uremic patients during hemodialytic treatment

We validated the Hypomania Interview Guide-Seasonal Affective Disorder version (HIGH-SAD) in patients with rapid cycling bipolar disorder (RCBD). Fourteen outpatients were rated on six separate occasions (total = 84 visits). On each visit the patients were rated with the HIGH-SAD and the Young Mania Rating Scale (YMRS) in a counterbalanced order. Clinical assessment was completed at the end of the visit by the treating psychiatrist. Patients were assessed as hypomanic/manic on 22 of the visits. Pearson correlation coefficient between the YMRS total scores and the HIGH-SAD total scores for those 22 visits in which patients were hypomanic/manic was r = 0.629 (P < 0.05) and for all visits was r = 0.769 (P < 0.0001). Analysis with only one rating per patient yielded a Pearson correlation coefficient of r = 0.792 (P < 0.0004). We found that the HIGH-SAD was a valid scale for the measurement of hypomania in patients with RCBD. However, the scale does not differentiate hypomania from mania in this group of patients.     

The robustness and applicability of speed–accuracy decomposition, a technique for measuring partial information.

Speed–accuracy decomposition (SAD) is a relatively new technique for studying the time course of information processing. It uses information on both the temporal distributions and the accuracies of participants' responses to 2 types of trials to derive an estimate of the amount of partial information available to participants at specific points in processing. As a new technique, its range of applicability and robustness have not yet been fully determined. Simulations are reported here that investigate these issues, with simulated data sets, SAD analysis of same, and comparison of the SAD analysis results to the parameters used to generate the data. In general, SAD appears to be reasonably robust against certain violations of assumptions and a variety of unusual data patterns. It is recommended that future SAD studies be accompanied by simulations based on observed parameters to better understand the data obtained. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Stability of nonlinear principal components analysis: An empirical study using the balanced bootstrap.

Principal components analysis (PCA) is used to explore the structure of data sets containing linearly related numeric variables. Alternatively, nonlinear PCA can handle possibly nonlinearly related numeric as well as nonnumeric variables. For linear PCA, the stability of its solution can be established under the assumption of multivariate normality. For nonlinear PCA, however, standard options for establishing stability are not provided. The authors use the nonparametric bootstrap procedure to assess the stability of nonlinear PCA results, applied to empirical data. They use confidence intervals for the variable transformations and confidence ellipses for the eigenvalues, the component loadings, and the person scores. They discuss the balanced version of the bootstrap, bias estimation, and Procrustes rotation. To provide a benchmark, the same bootstrap procedure is applied to linear PCA on the same data. On the basis of the results, the authors advise using at least 1,000 bootstrap samples, using Procrustes rotation on the bootstrap results, examining the bootstrap distributions along with the confidence regions, and merging categories with small marginal frequencies to reduce the variance of the bootstrap results. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Performance of the Norwegian SF-36 Health Survey in patients with rheumatoid arthritis. II. A comparison of the SF-36 with disease-specific measures

The performance of the SF-36 was compared with disease-specific health status instruments (Arthritis Impact Measurements Scales [AIMS2], Modified Health Assessment Questionnaire [MHAQ] and visual analogue scales) in 1030 patients with rheumatoid arthritis (mean age 62.3 years, 79% females, mean disease duration 12.9 years, 48% rheumatoid factor positive). The scales performed similarly in known group comparisons (age cohorts, disease severity, disease activity, comorbidity). The SF-36 physical functioning scale correlated -0.69 and -0.73 with the MHAQ and AIMS2 physical scales, respectively. A strong negative correlation was found with the walking and bending subscale of AIMS2 (r = -0.80), a substantial negative correlation with mobility (r = -0.65), and moderate correlations with the scales for hand/finger and arm function (r = -0.52 and r = -0.53). Frequency distributions of scores revealed more skewed distributions of the AIMS2 physical scale and the MHAQ scale than the physical functioning scale of the SF36, whereas the pain and mental health scales were distributed similarly. In conclusion, the SF-36 performs well in patients with rheumatoid arthritis. The physical functioning scale of the SF-36 does not seem to capture all aspects of physical health in rheumatoid arthritis patients, but may be more sensitive than disease-specific measures to low levels of physical disability.     

Alternatives to classical statistical procedures: Introduction to the symposium.

Most classical statistical methods are founded on two assumptions, that the underlying distribution is normally distributed, and that the sample observations are independently and identically distributed. It did not take statisticians long to realize that these idealizations may not be true in all applied settings. In addition, the good performance of these classical procedures requires strict adherence to these assumptions. Consequently, alternatives to the classical procedures have been proposed by statisticians. The purpose of this symposium is to begin a dialogue between quantitative methodologists and substantive psychological researchers (who are voracious consumers of the classical procedures) about alternatives to the classical methods. The first three papers (Hunter & May, Zumbo & Zimmerman, and May & Hunter) will discuss a class of methods referred to as nonparametric statistics. Nonparametric procedures arc an early and popular alternative to the classical procedures. Lind & Zumbo will discuss methods which are designed to deal with violations of normality. In closing, Thomas will discuss alternatives to classical statistical procedures when the second assumption is violated. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Mortality studies of machining fluid exposure in the automobile industry. V: A case-control study of pancreatic cancer

MANOVA and repeated measures ANOVA approaches have provided evidence of a number of limitations in several event-related potential (ERP) studies due to violations of their statistical assumptions and the typically moderate size of the available sample. Alternative, computer-intensive methods based on permutation principles have recently been developed. Up to now this methodology has focused mostly on magnitude differences between scalp distributions as measured by t statistics. In this paper the scope of permutation techniques in ERP analysis was widened. A new statistic (D statistic) is introduced to compare the shapes of scalp distributions of ERPs. Additionally a general non-parametric combinatory technique is introduced to evaluate, by means of multivariate permutation tests, several time points and/or recording sites in ERP data. The methodology described here was used to test if two ERP components elicited during word-pair matching tasks to semantic or phonological incongruences had different scalp distributions.     

The needless assumption of normality in Pearson's r.

With regard to proper application of the Pearson technique, statistical textbook authors fall into 2 groups: those who indicate that it is necessary for each of the variables to be normally distributed, and those who point out that normality of distribution is required in neither of the correlated variables. Perhaps the erroneous assumption of normally distributed variables stems "from Pearson's first presentation of the method and has been perpetuated through 11 Editions of Fisher's classic text. An objection to Pearson's choice of a poor example and his failure to generalize beyond the restrictive limits of normal data were pointed out at the time by Yule." Psychologists and statisticians "are still harassed in many quarters by the needless assumption of normality for the proper computation of Pearson's r." (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Identification of matrix metalloproteinases and metalloproteinase inhibitors in bovine corpora lutea and their variation during the estrous cycle

Corpora lutea (CL) were collected from cattle to study key physiologic events in angiogenesis. Our objective was to evaluate the activity of matrix metalloproteinases (MMP) and endogenous inhibitors. Corpora lutea were collected 2, 4, 6, 8, 10, 12, 14, and 16 d (n = 3/d) after estrus was first detected. In zymograms, a band of protein migrating at a relative molecular mass (M(r)) of 98 kDa was increased early in the cycle; a M(r) = 88 kDa band was detectable on all days. The molecular weights of these proteins are consistent with the MMP-9 family members. In all samples, a band of enzyme activity was detected at M(r) = 62 kDa, and another band of lesser density was detected at M(r) = 60 kDa. The molecular weights of these proteins are consistent with the MMP-2 family members. An immunoreactive band, detected in all samples with equal density, migrated between M(r) = 27 and 29 kDa, as did the tissue inhibitor of metalloproteinase (TIMP-1) standard. A second band, which was less dense in samples from d 2 through 6, migrated at M(r) = 19 kDa, as did the TIMP-2 standard. A third band was detected in all samples; it migrated at M(r) = 35 kDa, as did the cartilage-derived inhibitor (CDI) standard, and was less dense in d 8 and d 12 through 16 samples. In summary, MMP (gelatinases) and MMP inhibitors are present in developing luteal tissue, and the M(r) = 98 kDa enzyme, CDI, and TIMP-2 varied during the estrous cycle.