نشریه بررسی های آمار رسمی ایران
سال بیست و چهارم شماره 2 (پیاپی 83، پاییز و زمستان 1392)

Gaussian mixture model is the most widely used finite mixture model. The important feature of this model is its flexibility with respect to various forms of continuous distributions. Because the most important part of model fitting is the estimation of its parameters, we want to estimate two-component Gaussian mixture model parameters through the four estimation methods. We introduce a two-component Gaussian mixture model and then estimate model parameters using the method of moment and maximum likelihood by the name of analytical and numerical solutions, respectively. In continuation, the parameters are estimated using the EM algorithm and the Gibbs sampler algorithm. In conclusion, the results of these methods are compared with each other. We try to solve an estimation problem with the four common methods and introduce advantage and disadvantage of them for users.

In surveys from finite populations usually a probability sample is selected from a sampling frame containing all the units in the target population. In some cases the frame that is available cannot cover target population completely so survey statistics from these frames are bias. On the other hand updating the frame that is available is too expensive. Multiple frame survey can be used to increase population coverage, reduce bias and decrease costs. In multiple frame surveys samples are drown independently from overlapping multiple frames so that these frames give an appropriate coverage of the target population. Several estimators for population total in multiple frame surveys are discussed. In this article we introduce these estimators in a simulation study, we compare performance of different total estimators.

Questionnaire is one of the most popular statistical data collection tools therefore its design has a profound impact on quality of collected data. Applying principles like: avoiding ambiguous, complicated and general questions as well as choosing appropriate appearance and layout can improve questionnaire design quality. Assessment of these principles, particularly from content view point, may not be possible by looking at questionnaire only, so questionnaire evaluation before main survey implementation is suggested. Different methods have been introduced on questionnaire evaluation. This article presents "interviewer debriefing" method which is used in questionnaire evaluation of "Survey on producer price indices and production quantity on mining section-2009", Statistical Center of Iran, and the results has been analyzed. The results show that "interviewer debriefing" is a suitable method in questionnaire evaluation of this survey. In this article some recommendations to improve questionnaire design and filling out procedure are given to interviewers.

Unemployment rate is an indicator of socio-economic situation to allocate effectively the funds and conducting employment plans or policies we need a reliable estimate of employment and unemployment characteristic in different areas of country. The national statistical offices face to the problem that the sample size of current notional surveys are not planned to provide reliable direct estimates for such small areas and the increase in size that is necessary to cover all these areas adequately is not affordable. Thus we need to use the small area estimation techniques to make reliable estimate for these areas. In model-assisted approach that relate response variable to some auxiliary variables by a model, it is possible to improve precision of the small area estimators by borrowing strength from auxiliary variables. In this article logistic regression model and multinomial logit model with random area effect are introduced for estimate unemployment rate in ostans. Next in a simulation study by using 1389 labor force survey data we compare efficiency of model-based and design-based small area.

Hierarchical and K-mean clustering, according to their ability to meet the needs and problems of research in several scientific disciplines, are widely used among researchers. Extensive use of clustering in different contexts of health, engineering, social sciences and humanities, lead to growth of knowledge in these sciences. R, SAS and MATLAB software are the most widely used software in the analysis of clustering because of their ability in programming. However, each of this software has particular ability and limitations to clustering. So the familiarity of researcher with clustering method in each of this software allows them to use software which meets their needs for performing hierarchical clustering and K-means.

Youth population of country is the national capital and the largest source of development economic and cultural, hence detailed study trend of social indicators of the group and comparison with the total population trend indicators is the source of valuable information that help planners and policy makers to provide developing effective programs and policies.In this paper, demographic characteristics, migration, education and marital status among Iranian youth using the results of the Population and Housing censuses will be reviewed. In this paper will be studied, demographic characteristics, migration, education and marital status among Iranian youth by use the results of the Population and Housing censuses.Research results show changes in indices related to education and marriage among youth because of economic and social issues during the recent years.