Search results

Qualitative researchers have been criticised for not justifying sample size decisions in their research. This short paper addresses the issue of which sample sizes are appropriate and valid within different approaches to qualitative research.

The sparse literature on sample sizes in qualitative research is reviewed and discussed. This examination is informed by the personal experience of the author in terms of assessing, as an editor, reviewer comments as they relate to sample size in qualitative research. Also, the discussion is informed by the author’s own experience of undertaking commercial and academic qualitative research over the last 31 years.

In qualitative research, the determination of sample size is contextual and partially dependent upon the scientific paradigm under which investigation is taking place. For example, qualitative research which is oriented towards positivism, will require larger samples than in-depth qualitative research does, so that a representative picture of the whole population under review can be gained. Nonetheless, the paper also concludes that sample sizes involving one single case can be highly informative and meaningful as demonstrated in examples from management and medical research. Unique examples of research using a single sample or case but involving new areas or findings that are potentially highly relevant, can be worthy of publication. Theoretical saturation can also be useful as a guide in designing qualitative research, with practical research illustrating that samples of 12 may be cases where data saturation occurs among a relatively homogeneous population.

Sample sizes as low as one can be justified. Researchers and reviewers may find the discussion in this paper to be a useful guide to determining and critiquing sample size in qualitative research.

Sample size in qualitative research is always mentioned by reviewers of qualitative papers but discussion tends to be simplistic and relatively uninformed. The current paper draws attention to how sample sizes, at both ends of the size continuum, can be justified by researchers. This will also aid reviewers in their making of comments about the appropriateness of sample sizes in qualitative research.

To develop, test and implement a sampling strategy for equipment auditing for a Fortune 100 company.

Regression analysis is applied to auditing of equipment for a large US corporation. Empirical data and test data sets are used to evaluate the efficacy of using regression for auditing and to determine reasonable and efficient sample sizes to be employed across more than 5,000 locations.

Regression is a viable and useful method for equipment auditing when there is anticipated high correlation between pre‐ and post‐audit equipment value. Recommended sample size is dependent upon the size of the location as measured by total pieces of equipment. Decision rules combining acceptable tolerance limits, desired confidence level and sample size are provided.

The method, recommended sample sizes and decision rules are particularly applicable to instances where high correlation is expected between pre‐ and post‐audit equipment values. Standard regression assumptions are not all met in all instances, especially with small sample sizes.

The regression approach and model, sample size recommendations and decision rules for passing or failing an equipment audit described herein have been implemented at a Fortune 100 company, and are generally applicable to equipment and inventory auditing when high correlation between pre‐ and post‐audit equipment is expected.

This paper provides a practical and useful regression‐based approach to sampling for equipment auditing. Recommended sample sizes and decision rules for passing or failing the audit are explicitly defined.

The purpose of this paper is to examine the similarities and differences in human resource (HR) practices between large and mid‐sized organizations.

With organization life cycle as the theoretical base, this study collects data on HR practices from a sample of mid‐sized firms in one western state and compares the results to a previously published study of HR practices in much larger firms. Mid‐sized firms are defined as having between 300 and 3,000 employees (mean=1,205); the average number of employees in the large firm sample was 30,700.

While there are some similarities in HR practices, it appears that HR organizations in large firms are more targeted at strategic business issues and have achieved a greater degree of outsourcing and/or information technology application for the routine, file maintenance aspects of HR work. On the other hand, HR informants in mid‐sized firms are more operationally focused, appear to be more satisfied with the skills and knowledge of their significantly smaller HR staffs, and give their HR organizations higher ratings on performance along multiple dimensions.

Limitations to the present study include the use of a single informant and a limited sample of mid‐sized firms (n=32). However, the study suggests numerous possibilities for future research, including expansion of the sample both in size and geographic scope, the collection of measures of performance for comparison with those of HR satisfaction, and fuller development of a set of expectations for HR focus dependent on the size and rate of growth of the organization.

The HR practices of large and small (less than 500 employees) firms have been studied in the past. This paper fills a gap in organization life cycle theory by examining HR practices in mid‐sized firms, suggesting in particular that the drive to refocus HR on strategic issues in large firms may be inappropriate in enterprises with 300 to 3,000 employees.

The authors derive the necessary mathematics, provide computer simulations, provide links to free and user-friendly computer programs, and analyze real data sets.

Cohen's d, which indexes the difference in means in standard deviation units, is the most popular effect size measure in the social sciences and economics. Not surprisingly, researchers have developed statistical procedures for estimating sample sizes needed to have a desirable probability of rejecting the null hypothesis given assumed values for Cohen's d, or for estimating sample sizes needed to have a desirable probability of obtaining a confidence interval of a specified width. However, for researchers interested in using the sample Cohen's d to estimate the population value, these are insufficient. Therefore, it would be useful to have a procedure for obtaining sample sizes needed to be confident that the sample. Cohen's d to be obtained is close to the population parameter the researcher wishes to estimate, an expansion of the a priori procedure (APP). The authors derive the necessary mathematics, provide computer simulations and links to free and user-friendly computer programs, and analyze real data sets for illustration of our main results.

In this paper, the authors answered the following two questions: The precision question: How close do I want my sample Cohen's d to be to the population value? The confidence question: What probability do I want to have of being within the specified distance?

To the best of the authors’ knowledge, this is the first paper for estimating Cohen's effect size, using the APP method. It is convenient for researchers and practitioners to use the online computing packages.

Addresses the standardization of the measurements and the labels for concepts commonly used in the study of work organizations. As a reference handbook and research tool, seeks to improve measurement in the study of work organizations and to facilitate the teaching of introductory courses in this subject. Focuses solely on work organizations, that is, social systems in which members work for money. Defines measurement and distinguishes four levels: nominal, ordinal, interval and ratio. Selects specific measures on the basis of quality, diversity, simplicity and availability and evaluates each measure for its validity and reliability. Employs a set of 38 concepts ‐ ranging from “absenteeism” to “turnover” as the handbook’s frame of reference. Concludes by reviewing organizational measurement over the past 30 years and recommending future measurement reseach.

The purpose of this paper is to form a numerical simulation method for permeability coefficient that can consider the characteristics of gravel gradation and further explore the effects of indoor test factors and gradation characteristics on the permeability coefficient of gravel.

The random point method is used to establish the polyhedral gravel particle model, the discrete element method (DEM) is used to construct the gravel permeability test sample with gradation characteristics and the finite element method is used to calculate the permeability coefficient to form a DEM-computational fluid dynamics combined method to simulate the gravel seepage characteristics. Then, verified by the indoor test results. Based on this method, the influence of sample size, treatment method of oversize particles and the content of fine particles on the permeability coefficient of gravel is studied.

For the gravel containing large particles, the larger size permeameter should be used as far as possible. When the permeameter size is limited, the equal weight substitution method is recommended for the treatment method of oversized particles. Compared with the porosity, the pore connectivity has a higher correlation with the permeability coefficient of the sample.

Insufficient consideration of the movement of gravel particles in the seepage process is also an issue for further study.

The simulation method described in this paper is helpful for qualitative analysis, quantitative expression of pore size and makes up for the defect that the seepage characteristics in pores cannot be observed in laboratory tests.

It is found that the process outputs from most of the industries are correlated and the performance of X-bar chart deteriorates when the level of correlation increases. The purpose of this paper is to compute the level of correlation among the observations of the weights of tablets of a pharmaceutical industry by using modified X-bar chart.

The design of the modified X-bar chart is based upon the sum of χ2s, using warning limits and the performance of the chart is measured in terms of average run lengths (ARLs). The ARLs at various sets of parameters of the modified X-bar chart are computed; using MATLAB software at the given mean and standard deviation.

The performance of the modified X-bar chart is computed for sample sizes of four. ARLs of optimal schemes of X-bar chart for sample size of four are computed. Various optimal schemes of modified X-bar chart for sample size (n) of four at the levels of correlation (Φ) of 0.00, 0.25, 0.50, 0.75 and 1.00 are presented in this paper. Samples of weights of the tablets are taken from a pharmaceutical industry and computed the level of correlation among the observations of the weights of the tablets. It is found that the observations are closely resembled with the simulated observations for the level of correlation of 0.75 in this case study. The performance of modified X-bar chart for sample size (n) of four at the levels of correlation (Φ) of 0.50 and 0.75 is also compared with the conventional (Shewhart) X-bar chart and it is concluded that the modified X-bar chart performs better than Shewhart X-bar chart.

All the schemes are optimized by assuming the normal distribution. But this assumption may also be relaxed to design theses schemes for autocorrelated data. The optimal schemes for modified X-bar chart can also be used for other industries; where the manufacturing time of products is small. This scheme may also be used for any sample sizes suitable for the industries

The optimal scheme of modified X-bar chart for sample size (n) of four is used according to the computed level of correlation in the observations. The simple design of modified X-bar chart makes it more useful at the shop floor level for many industries where correlation exists. The correlation among the process outputs of any industry can be find out and corresponding to that level of correlation, the suggested control chart parameters can be used.

The design of modified X-bar chart uses very less numbers of parameters so it can be used at the shop floor level with ease. The rejection level of products in the industries can be reduced by designing the better control chart schemes which will also reduce the loss to the society as suggested by Taguchi (1985).

Although; it is the extension of previous work but it can be applied to various manufacturing and service industries; where the data are correlated and normally distributed.

This study investigated changes in appearance due to variations in the amount of ease allowance at the bust, waist and hips with ready-made women's shirts in two different styles. The authors also examined the suitable range for ease allowance with those two styles using a sensory test for evaluating appearance.

The authors employed two women's shirts (samples I and II) in different styles (fit and straight). The authors modified the size of the bust, waist and hips with a changeable-size dress form using 1-cm intervals. The authors observed the shape changes. With the pictures at 2-cm intervals, the authors conducted a sensory test to evaluate appearance. 20 Japanese university students in their 20s made their assessments using seven items (wrinkles, fit, silhouette, beauty, fashionable impression, comfort and purchase intention) with a five-point scale. The proportion of subjects who scored 1 or more was 40% or greater when using ease allowance in the suitable range.

The appropriate ranges of ease allowance differed according to the style and evaluation items. Regarding appearance and purchase intention with sample I from the front, the suitable range of ease allowance was 1–7 cm for the bust, 2–6 cm for the hips and 13 cm for the waist. From the side, the range was 0–6 cm for the hips and 7 cm for the bust. With sample II from the front, the suitable range of ease allowance was about 4 cm for all parts. From the side, the range was 11 cm for the bust and 4 cm for the hips.

Using a dress form, the authors determined suitable ranges of ease allowance for two women's shirts in fit and straight styles with seven bust, six waist and six hip sizes. The authors found that the suitable range of ease allowance varied according to the style, direction and parts of the shirt. Our results can act as a guideline for designing and selecting ease allowance for women's shirts, taking into account comfort from appearance, purchase intention and beauty.

The purpose of this paper is to provide a rigorous, statistically correct, and low‐cost way to audit sample a lender's loan portfolio, be they a microlender or other type of lender. No other paper applies this method to loan portfolios, even though it is a high demand application.

Standard techniques of audit sampling and dollar unit sampling with stratification are applied to the particular case of a microlender's portfolio. Unlike the audit sampling that almost all auditors use, no arbitrary rules of thumb are applied.

The paper finds that statistical audit sampling for a lender's loan portfolio is simple, rigorous, and inexpensive.

In audit sampling, most auditors use arbitrary rules of thumb and have no idea whether they are sampling enough items to actually be sure, with some desired level of confidence, that they have found no defects. This simple, inexpensive, and statistically rigorous technique will allow auditors who actually want to do a good job to quantify the precision of their statements in a very common application.

This paper combines several disparate threads from the statistical literature on audit sampling in a way that auditors (who are usually not statisticians) can apply them for auditing the quality of a lender's portfolio – microfinance or otherwise – which is a very common need.

Bayesian networks (BNs) are implemented for monitoring a process via statistical process control (SPC) where attribute data are available on output from the system. The paper aims to discuss this issue.

The BN provides a graphical and numerical tool to help a manager understand the effect of sample observations on the probability that the process is out-of-control and requires investigation. The parameters for the BN SPC model are statistically designed to minimize the out-of-control average run length (ARL) of the process at a specified in-control ARL and sample size.

The BN model outperforms adaptive np control charts in all experiments, except for some cases where only a large change in the proportion of sample defects is relevant. The BN is particularly useful when small sample sizes are available and when managers need to detect small changes in the proportion of defects produced by the process.

The BN model is statistically designed and parameters are chosen to minimize out-of-control ARL. Future advancements will address the economic design of BNs for SPC with attribute data.

The BNs allow qualitative knowledge to be combined with sample data, and the average percentage of defects can be modeled as a continuous random variable. The framework of the BN easily permits classification of the system operation into two or more states, so diagnostic analysis can be performed simultaneously with statistical inference.