Sampling designs

An email list of 5,000 target respondents were purchased from Council of Supply Chain Management Professionals (CSCMP), a prestigious association of professionals in the area of supply chain management from diverse industries 

The email list was relined to eliminate multiple names from the same organization. The person with the most relevant job title was picked and the others were removed. In this process, 249 names were removed from the email list. An invitation to participate in the survey, which explained the purpose of study, the instructions for completing the questionnaire, and measures to securely handle the data collected, were sent via email to 4,751 potential respondents. For the convenience of the respondents, three options were provided to complete and submit the questionnaire (1) On-line Click on the link that would take the respondents to the survey website to complete the survey (2) Fax Click on the link that would allow the respondents to download a copy of the questionnaire and send it by fax (3) Regular mail Email back to request a hard copy to be sent through regular mail and return it through either fax or regular mail. 

To improve the response rate, three waves of emails were sent once a week. A total of 152 responses were obtained on-line after the first wave of emails. The second and third wave generated 71 (69 on-line, 2 via mail) and 4 (2 on-line, 2 via mail) responses respectively. Out of the 227 responses received (16 incomplete), 211 are usable resulting in a response rate of 6.0 % (211/3538). Based on the information collected on the website, the number of unique clicks (one click per IP address is counted) is 702 resulting in a click through rate of 19.8 % (702/3538). The response rate out of the unique clicks is 30.1 % (211/702). 

Variables First-wave frequency Second/third wave (expected frequency fe) Second/third wave (observed frequency f ) Chi-square test 

Sample characteristics appear on Table 7.2 based on SIC code, firm size, and respondents job titles. The respondents come from manufacturing industries, namely, SIC 25, 30, 34, 35, 36, 37 and 38. The highest four respondent categories by SIC code are 34, 35, 36, and 37 (i.e., 79 % of respondents). About 80 % of firms have 100-500 or 1001 and more employees. 73 % of the respondents are presidents/CEO vice presidents 24 % are managers and directors. 

---

Sample designation Melt flow rate, dg/min Stiffness, MPa Tensile properties ... 

The need for skill and experience on the part of sample designers and persoimel cannot be overemphasized in chemical plant sampling. Safety precautions are of the utmost importance. Necessary steps must be taken to document the hazards involved in an operation and to ensure that the staff are weU-trained, informed, protected, and capable. Except for bulk powder sampling, most chemical plant sampling is hazardous and difficult and must be designed with care. The following discussions are based on the assumptions that most of these decisions have been made and a satisfactory sampling procedure has been planned. 

R. G. Nash and A. R. LesHe, eds.. Groundwater Residue Sampling Design, American Chemical Society, Washington, D.C., 1991. 

Sampling spills. Wear proper personal protective equipment (PPE) Follow proper sampling procedures Use safe sampling design CCPS G-22 CCPS G-23 CCPS G-29 Lovelace 1979... 

In this subsection, basic design theory for preliminary sizing and specifying equipment are reviewed. Some sample design calculations are included. References cited at the end of tlie chapter can be consulted for more detailed information and design methods. For solid-liquid separation methods, the reader should refer to Liquid Filtration, 2" edition, by N. P. Cheremisinoff, Butterworth-Heinemarui Publishers (1998). 

Table 4. Design specification for pressure relief valves (sample design specification format).

The blends prepared by twin-screw extruder with two different draw ratios were extruded with the Brabender single-screw extruder at temperatures ranging from 180 280°C, The sample designation and specific processing conditions are given in Table 1. 

A comprehensive theory encompassing sampling design through final error analysis that uses the spatial or temporal correlation of environmental samples to optimize sampling and analysis. 

Many characteristics of Interest In an environmental sampling study require chemical or physical analysis In a laboratory. When a laboratory Is Involved, the sampling design must Include consideration of sample processing In the laboratory and analytical protocols. 

Determine the components of variance that should be built Into the statistical design Proper stratification of the study area will allow Identification and quantification of several sources of variation. The sources of variation that can be controlled by the sampling are determined by the particular sampling design and by the pattern of sample collection superimposed over the area. An analysis of variance of the data provides estimates of the components of variance ... 

Danz NP, Regal RR, Niemi GJ, Brady VJ, Hollenhorst T, Johnson LB, Host GE, Hanowski JM, Johnston CA, Brown T, Kingston J, Kelly JR. 2005. Environmentally stratified sampling design for the development of Great Lakes environmental indicators. Environ Monitor Assess 102 41-65. 

Protocols must specify the number of study sites to be sampled for residues and the number of stations within a study site to be sampled. This is a difficult process to specify, but depends on traditional sampling theory. In general, the parameters that need to be evaluated for proper sampling design are the following ... 

Sampling techniques must be defined and controlled. If leaf punch samples are taken, the sampling design must define the portions of the crop that will be sampled (e.g., all sides of each plant, how high to sample in trees, inner and outer leaves). Never sample a leaf by brushing with an outer leaf that may be sampled at a later date. 

For tree crops, the EPA recommends the Iwata approach. In this approach, 40 punches are collected for each sample at various heights and at 45° intervals around the circumference of each sampled tree. The sampling design and technique are described below. 

The identification of sampling requirements involves specifying the sampling design, the sampling method, sample numbers, types, and locations, and the level of sampling quality control. Data quality requirements include precision, accuracy, representativeness, completeness, and comparability. 

The sampling design allowed for excellent acceptance by operators, maximized performance of the analytical method, and provided clear, conclusive results with minimal numbers of different sample types. 

The methodology used for each of the case studies presented in the next section shows some differences that are summarized in Table 1. In most cases, the data were collected following a random sampling design. Only in case V, the sampling was systematic. 

In the case that inhomogeneities are not attributed to random variations but to systematic changes as, e.g. shown in Fig. 2.5B, then other strategies of sampling and evaluation must be used. Numerous different sampling designs have been proposed. Surveys can be found, e.g. in Cochran [1977], Gilbert [1987], Keith [1988] Nothbaum et al. [1994], and Einax et al. 

The direct calculation of the reduced partition function ((e-AU kT))o,h=i is expected to be limited by the variance of the function e-AU/kT [ Iowever, representing the averaged quantity as a ratio opens possibilities for importance sampling, evaluating both numerator and denominator on the basis of a single sample designed to reduce the variance of e AU/kTm... 

To improve the processing properties, the material was ground in a ball-mill for two hours (sample designated as 25B2m2) and separately for 6 hours (sample designated as 25B2m6). In order to check the effect of milling the particle distribution of all the samples was determined (see Fig. 4). 

A relatively large number of indoor radon measurements have already been made, particularly during the last 6-12 months. However, most of these measurements are not useful in defining the national frequency distribution of indoor radon levels because there are a number of uncertainties and limitations associated with these measurements. These uncertainties include the different purposes for the measurements, the widely varying sample designs, and the many different sample collection and measurement procedures used. 

Alkaline earth oxides (AEO = MgO, CaO, and SrO) doped with 5 mol% Nd203 were synthesised by both evaporation and sol-gel methods. According to the first method, water solution of nitrates was evaporated under continuous stirring, dried, and nitrate mixture was decomposed at corresponding temperatures (Table 1) for one hour (samples designated Ev). All samples were calcined at 650°C for 3h after the decomposition. In the sol-gel method the samples (designated SG) were obtained... 

Table 2. Soil sampling design of the selected soil profiles at the Dinghushan Biosphere Reserve, South China ...

FAO classification (FAO 1998). Soil sampling design and the number of samples collected from each profile are provided in Table 2. 

---