Variable-response pairs

Scatter plots are used to indicate reiationships between two variables or pairs of data (Figure 16-12). The easiest way to determine if a cause-and-effect relationship exists between two variables is to use a scatterpiot. Variables such as flow rate and temperature can be used in a scatter diagram. In this relationship, temperature may increase or decrease as flow rate fluctuates. This response would show up on a scatterpiot diagram on the x- or y-axis. The independent variable is typically controllable, while the dependent variable is located on the opposite axis. 

The term s will refer to the random error associated with the variable response, and, finally, the model,/(x), is a mathematical function that relates y to X. It is a working hypothesis and must be modified if the experimental data are against it. In analytical chemistry we assume some sort of cause (level of the property)-and-effect (signal variation) relation and, hence, it is reasonable to accept that the model for the observed experimental responses isy=/(x)+ , where f(x) is the standardization (calibration) curve to be estimated from the experimental data points. Of course, for each measurement we can state yi=f(Xi) + , i= 1,.., N, where x y, are the data pairs associated with each calibrator. 

Since only 20 data records were collected from the system during the execution of the designed experiments conducted by Reece et al. (1989), we used their response surface models, deliberately contaminated with small Gaussian noise terms, to generate a total of 500 (x, z) pairs (assuming that the three variables, jCj, Xj, Xj, have independent and uniform... 

In the calibration problem two related quantities, X and Y, are investigated where Y, the response variable, is relatively easy to measure while X, the amount or concentration variable, is relatively difficult to measure in terms of cost or effort Furthermore, the measurement error for X is small compared with that of Y The experimenter observes a calibration set of N pairs of values (x, y ), i l,...,N, of the quantities X and Y, x being the known standard amount or concentration values and y the chromatographic response from the known standard The calibration graph is determined from this set of calibration samples using regression techniques Additional values of the dependent variable Y, say y., j l,, M, where M is arbitrary, are also observed whose corresponding X values, x. are the unknown quantities of interest The statistical literature on the calibration problem considers the estimation of these unknown values, x, from the observed and the... 

In the paired t-test setting it is the normality of the differences (response on A — response on B) that is required for the validity of the test. The log transformation on the original data can sometimes be effective in this case in recovering normality for these differences. In other settings, such as ANOVA, ANCOVA and regression, log transforming the outcome variable is always worth trying, where this is a strictly positive quantity, as an initial attempt to recover normality. 

A useful computer programme available as a supplement to regression is known as the response surface plot This programme calculates the values of the independent variables which combine to give a specific value of the response) and it plots contours of the specific values on coordinates made up of pairs of the independent variables. 

Are there relationships among the dependent or response variables By rearranging the data, the correlations between ell possible pairs may be evaluated. 

It appears that the formal theories are not sufficiently sensitive to structure to be of much help in dealing with linear viscoelastic response Williams analysis is the most complete theory available, and yet even here a dimensional analysis is required to find a form for the pair correlation function. Moreover, molecular weight dependence in the resulting viscosity expression [Eq. (6.11)] is much too weak to represent behavior even at moderate concentrations. Williams suggests that the combination of variables in Eq. (6.11) may furnish theoretical support correlations of the form tj0 = f c rjj) at moderate concentrations (cf. Section 5). However the weakness of the predicted dependence compared to experiment and the somewhat arbitrary nature of the dimensional analysis makes the suggestion rather questionable. 

How do we set about variable selection One obvious approach is to examine the pair-wise correlations between the response and the physicochemical descriptors. One form of model building, forward stepping multiple regression, begins by choosing the descriptor that has the highest correlation with a response variable. If the response is a categorical variable such as toxic/non-toxic,... 

---