Sampling differences between studies

While we can find the optimal parameters for any given frame for a model such as LF, we never actually achieve a perfect fit. In fact, we can determine a sample by sample difference between the model and the residual. This of course, is another residual or error signal, which now represents the modelling error after the LP model and the LF model have been taken into account. One interesting study by Plumpe et al [356] describes how this signal is in fact highly individual to speakers and can be used as the basis of speaker recognition or verification. 

A typical example, from the extensive study by Kamakin on an alumina-silica gel, is shown in Fig. 3.32. When the mercury pressure was reduced to 1 atm at the end of the first cycle, 27 per cent of the intruded mercury was retained by the sample a second intrusion run followed a different path from the first, whereas the second extrusion curve agreed closely with the first. Change in f re structure of the kind described above could perhaps account for the difference between the two intrusion curves, but could not explain the reproducibility of the remainder of the loop. There is no doubt that hysteresis can exist in the absence of structural change. 

Our objectives in this section are twofold to describe and analyze a mechanical model for a viscoelastic material, and to describe and interpret an experimental procedure used to study polymer samples. We shall begin with the model and then proceed to relate the two. Pay attention to the difference between the model and the actual observed behavior. 

There is an important difference between the two techniques in that photons, produced by XRF, can pass through a relatively large thickness of a solid sample, typically 4000 nm, whereas electrons can penetrate only about 2 nm. This means that AES is more useful in the study of solid surfaces, whereas XRF gives information referring more to the bulk of a solid or liquid. 

In this study, four Styragel columns were utilized one column had a nominal porosity rating of 10, two colvtmns of 10, and the fourth column of 10 A. The refractometer was maintained at 37°C. A 5 ml syphon was used to monitor a solvent flow rate of 1 ml/min. The instrviment was run at the highest sensitivity setting because the refractive index difference between our solvent and polymer was only moderate and because a number of samples analyzed had a broad molecular weight distribution (MWD). 

This paper discusses the role that statistics can play In environmental sampling. The primary difference between an Investigation based on statistical considerations and one that Is not Is the degree of objectivity that can be Incorporated Into the evaluation of the quality and uncertainty of the study results. Statistical methods In the planning stage can also aid In optimizing allocation of resources. 

---