Standard linear regression

Instrument calibration is done during the analysis of samples by interspersing standards among the samples. Following completion of the samples and standards, a linear calibration curve is estimated from the response of the standards using standard linear regression techniques. The calibration constants obtained from each run are used only for the samples quantitated in that run. Drastic changes or lack of linearity may indicate a problem with the detector. 

An equation representing area versus concentration was determined using a standard linear regression analysis applied to the injection standards, yielding a slope m and an intercept b. The following equation was then used to calculate the concentration of the sample injected from the area measured ... 

The above linear regression problem can be readily solved using any standard linear regression package. 

Our approach to the first strategy requires that we construct a table with the pairs of values of x, and yt fisted in order of increasing values of T, (percentage response). Beside each of these columns a set of blank columns should be left so that the transformed values may be fisted. We then simply add the columns described in the linear regression procedure. Log and probit values may be taken from any of a number of sets of tables and the rest of the table is then developed from these transformed x and j/- values (denoted as x and y ). A standard linear regression is then performed. 

The power law viscosity model was developed by Ostwald [28] and de Waele [29]. The model has been used in the previous sections of this chapter and it has the form shown in Eq. 3.66. The model works well for resins and processes where the shear rate range of interest is in the shear-thinning domain and the log(ri) is linear with the log (7 ). Standard linear regression analysis is often used to relate the log... 

Take, for example, Anscombe s Quartet data set shown in Figure 20." Here, there are four sets of 11 pairs of XY data which when subjected to the standard linear regression approach all yield identical numerical outputs. Cursory inspection of the numerical listing does not reveal the information immediately apparent when simply plotting the XY data (Figure 21). The Chinese proverb that a picture is worth ten thousand words is amply illustrated. 

If, by reason of difficulty or lack of time, problems of weighting are to be ignored, all factors Wj can be set equal to 1 in the equations that follow. In that case however, subsequent sections dealing with evaluation of uncertainties and goodness of fit lose much of their validity unless applied to a case in which all the observations just happen to be of equal weight. The assumption w,- = 1 is implicit in standard linear regression operations of spreadsheet programs. 

A value of 5.2A for the ion-size parameter a yielded straight-line plots of E0 vs. m at each temperature and solvent composition. The intercepts were obtained by standard linear regression techniques. A graphical representation of the data for each of the H20/NMA solvent mixtures at 25°C is shown in Figure 1. The calculations were performed with the aid of a PDP-11 computer with a teletype output. The intercepts (E°) and the standard deviations of the intercepts are summarized in Table III. 

Eq. (6.12) describes a non-linear model. The term non-linear is used here, and only in this section, in a statistical sense. In that sense an equation such as Eq. (6.2) is a linear regression model although it describes a quadratic, and therefore curved, relationship. It is however linear in the h parameters and standard linear regression techniques can be applied to obtain them. In Eq. (6.12) the parameters are part of the exponent and transformations such as the log transform cannot help that. The regression model is then called non-linear. Non-linear regression is less evident than linear regression. Software is available but it turns out that non-linear regression can lead to unstable numerical results. How to avoid this is described in Ref. [69]. 

Standard linear regression and method of residual analyses of two-compartment IV infusion (zero-order absorption) data is limited to samples collected during the postinfusion period. Plasma concentrations during the infusion period do not lend themselves to a linear analysis for a two-compartment model. Estimation of parameters from measured postinfusion plasma samples is quite similar to the two-compartment bolus IV (instantaneous absorption) case. Proper parameter evaluation ideally requires at least three to five plasma samples be collected during the distribution phase, and five to seven samples be collected during the elimination phase. Area under the curve (AUC) calculations can also be used in evaluating some of the model parameters. 

We perform a standard linear regression and find very high correlation (Table 3.11). 

Fig. 2. Fibrin zymography of PA standards. (A) Fibrin zymogram demonstrating electrophoretic migration of uPA (45-kDa), tPA (70-kDa), and higher molecular weight tPA/PAI-1 complex (110-kDa). Fibrin indicator gels were stained with amido black (see Section 3.5). Samples 1-5 correspond to PAs and PAIs of various compositions as described (R4, R5). (B) Calibration curve obtained with PA standards. Linear regression analysis was performed and correlation coefficient (r) is shown (r = 1.000, perfect correlation).

Notice again that the k s occur linearly no matter what the functions are, and so standard linear regression methods, including various weighting, and so on. 

The difference between this and "standard" linear regression is that Variable 1 and Variable 2 can be any functions of x and y, as long as they are not combined in any way (i.e. you can t have ln(x + j) as one variable). The technique can be extended to more than two variables using a method called w multiple linear regression but as that s more difficult to perform, this section will focus on two-dimensional regression. 

The constants oq, oi, ctm are estimated by a standard linear regression technique i.e., by minimizing ... 

Using the reference standards, linear regression of SCy on SSy yields parameters (slope, K, and intercept, b) for calculating SCy in the unknowns. Parameters K and b from the regression can then be used to calculate SCy in the unknowns by... 

In weighted, least-squares analysis, it is assumed that the variance of the individual data points may be variable. In order to reduce this problem to the standard linear regression framework, a weight, w is introduced for each observation that reflects how good the data point is. Thus, the regression model for weighted, least-squares is... 

All prediction error models can be fit using standard, linear regression. 

Thus, a plot of the natural logarithm of 1 — [EI ]/[Et-] as a function of time should yield a straight line (Fig. 5.1 ). The slope of the line, which corresponds to —k, can be determined using standard linear regression procedures. 

To obtain an estimate of a k versus [Iq] data set has to be created at a fixed substrate concentration. A plot of this k versus [Iq] data would yield a straight hne (Fig. 5.2). With the aid of standard linear regression procedures, the value of the slope of this line can be obtained. This slope corresponds to... 

---