Edward R. Mansfield

Effects of Carbohydrate-Electrolyte Content of Beverages on Voluntary Hydration in a Simulated Industrial Environment

This study examined the effects of ingesting beverages of varying electrolyte-carbohydrate (ECHO) composition on hydration, sensory response, physiological function, and work performance during 4 hours of simulated industrial work for subjects wearing impermeable protective clothing (PC). Male subjects (N=18) completed four separate work sessions. Each session consisted of 30 min of treadmill walking with intermittent arm curls at 300 kcal per hour (moderate work rate), followed by 30 min of rest, for a total of 4 hours at 33 degrees C wet-bulb globe temperature. Excessive physiological strain prevented only four subjects from completing the 4-hour protocol. A different beverage was provided for consumption ad libitum for each work trial in a repeated measures, double-blind design. The beverages included lime colored water (W), lemon-lime placebo (P), lemon-lime ECHO with 18 mEq/L NaCl (ECHO18), and lemon-lime ECHO with 36 mEq/L NaCl (ECHO36). There was no difference in sweat production among the four trials (p = 0.61). Mean (standard deviation [SD]) fluid consumption was significantly greater for the ECHO36 [771 (+/-264) mL per hour] as compared with the W [630.6 (+/-234) mL per hour] and the P [655.2 (+/-228) mL per hour] (p<0.05), but not significantly greater than the ECHO18 [740.4 (+/-198) mL per hour]. Also, consumption of the ECHO18 was significantly greater than the W. Mean (SD) weight change, expressed as a percentage of total body weight (pre minus post), was -0.55(+/-0.8) for W, -0.31(+/-1.0) for P, -0.01(+/-1.1) for ECHO18, and +0.11(+/-1.1) for ECHO36 (p = 0.06). Subjects drank less and tended to experience greater weight loss in trials in which W or P were provided compared with trials in which either ECHO was provided. Thus, ECHO beverages, when provided ad libitum to workers wearing PC in a hot environment, produced better hydration than water.

An Approximate Test for Comparing Heteroscedastic Regression Models

Abstract This article addresses the problem of testing whether the vectors of regression coefficients are equal for two independent regression models when the error variances are unequal. The usual Chow statistic, appropriate when equality of variances can be assumed, is modified by replacing the pooled residual variance in the denominator with a weighted average of the residual variances from each data set. The weights are functions of the mean of the eigenvalues of W = X′ 1 X 1(X′ 1 X 1 + X′ 2 X 2)-1. Both numerator and denominator are then approximated by scalar multiples of chi-squared distributions. The parameters of these approximated distributions are chosen to equate their first two moments to those of the exact distribution. The resulting approximation for the modified Chow statistic, C*, is an F distribution with degrees of freedom that depend on the two sample sizes, the number of regressor variables, the average eigenvalue of W, and the true ratio of error variances. Since the latter is unknow...

Body composition prediction in university football players.

This study was intended to determine if previously-developed body composition prediction equations were valid for use with a Division I university football team. A sample of 68 Division I football players with a mean age of 19.7 yr, was assessed for body density (BD) by underwater weighing (UWW), residual volume by helium dilution, and 26 selected anthropometric measures. A predicted BD was obtained by using two sets of equations developed from college football players and from three generalized equations. The differences between predicted and observed body densities were analyzed. Seven of the nine models examined failed to accurately predict the BD for this population of university football players. One sport-specific equation of White, Mayhew, and Piper for individuals in the backfield and a generalized model of Jackson and Pollock (JP) containing two circumferences performed well when considering the mean of differences and the magnitude of total error relative to the published standard error. However, both of these models overestimate body density for players with low BD and underestimate BD when actual BD is high. Using the JP model for a player whose actual BD is near the sample mean of 1.070, the estimated mean is very close at 1.069. However, for players with actual BD of 1.050, the estimated mean is 1.054, and if actual BD is 1.085, the JP estimated mean is 1.078. The bias is linear between these points.

The Impact of Emergent Network Structure on Organizational Socialization

Network analysis allows researchers to study patterns of roles or social relationships within organizations, and is especially usefulfor examining the impacts of informal systems on organizational socialization. This article presents a study of the level of connections within networks based on information flow, affect, influence, and the exchange of goods or services. Controlling for belief system, congregation size, and size of community, researchers selected 44 protestant churches for the sample and administered questionnaires randomly to congregation members. Researchers used the resulting data to analyze the effects of the four types of networks on knowledge, desire for growth, personal development, attendance rates, and outreach. The findings indicate that informal systems greatly affect socialization, especially when network connections are based on information flow. The authors conclude that further research should address additional network characteristics that may influence organizational socialization, should employ network analysis at the micro level, and should examine the actual process or organizational socialization.

Diagnostic Value of Residual and Partial Residual Plots

Abstract This article illustrates why partial residual plots in addition to the usual residual plots are useful in a multiple regression analysis. The expected values of the vector of residuals and the vector of partial residuals are presented and examined for the situation when a regressor variable is misspecified in the model. If curvature exists in a predictor variable, the plot of residuals versus the variable displays the points scattered around a line that is a linear transformation of the correct functional form of the variable. Hence a nonrandom pattern may appear in the plot, but the appropriate transformation may not be evident. For a partial residual plot, the underlying signal displays the correct functional form of the predictor variables across the relevant range of interest, except in instances when severe collinearity exists.

Validation of body composition models for high school wrestlers

This study investigates the utility of two equations for predicting minimum wrestling weight and three equations for predicting body density for the population of high school wrestlers. A sample of 54 wrestlers was assessed for body density by underwater weighing, residual volume by helium dilution, and selected anthropometric measures. The differences between observed and predicted responses were analyzed for the five models. Four statistical tests were used to validate the equations, including tests for the mean of differences, proportion of positive differences, equality of standard errors from regression, and equivalence of regression coefficients between original and second sample data. The Michael and Katch equation and two Forsyth and Sinning equations (FS1 and FS21) for body density did not predict as well as expected. The Michael and Katch equation tends to overpredict body density while FS1 underpredicts. The FS2 equation, consisting of a constant adjustment to FS1, predicts well near the mean but not at the ends of the sample range. The two Tcheng and Tipton equations produce estimates which slightly but consistently overpredict minimum wrestling weight, the long form equation by 2.5 pounds and the short form by 3.8 pounds. As a result the proportion of positive differences is less than would be expected. But based on the tests for the standard errors and regression coefficients, the evidence does not uniformly reject these two equations.

An approximate test for comparing independent regression models with unequal error variances

Abstract The usual F statistic for comparing two independent regression equations is commonly used by practitioners. This test, however, presumes the equality of the error variance of the two populations. For applications where this assumption is not valid, an approximate test based on the same statistic is proposed that improves the Toyoda (1974) approximation by using Satterthwaites (1946) approximation not just for the numerator but also for the denominator of the usual F statistic. The unconditional significance level of this approximate test is computed for a variety of design configurations. The power of the approximate test relative to an upper bound is also considered.

TEACHING THE STANDARD DEVIATIONS OF LINEAR COMBINATIONS OF RANDOM VARIABLES USING GRAPHICS

ABSTRACT This article gives a geometric interpretation that views the standard deviation of a linear combination of random variables in terms of the length of the hypotenuse of a triangle and its calculation as equivalent to the theorem of Pythagoras. This illustration shows students why the coefficients are squared and the variances are added even when one or more of the coefficients are negative. Linear combinations of two variables are discussed for both the independent and the dependent situations, and the extension to three variables is illustrated for the independent case.