Sampling procedure
<p style="border:solid thin black;"> THE CLEANED AND HARMONIZED VERSION OF THE SURVEY DATA PRODUCED AND PUBLISHED BY THE ECONOMIC RESEARCH FORUM REPRESENTS 100% OF THE ORIGINAL SURVEY DATA COLLECTED BY THE DEPARTMENT OF STATISTICS OF THE HASHEMITE KINGDOM OF JORDAN</p>
Survey Frame
=========
The sample of this survey is based on the frame provided by the data of the Population and Housing Census, 2004. The Kingdom was divided into strata, where each city with a population of 100,000 persons or more was considered as a large city.
The total number of these cities is 6. Each governorate (except for the 6 large cities) was divided into rural and urban areas. The rest of the urban areas in each governorate was considered as an independent stratum. The same was applied to rural areas where it was considered as an independent stratum. The total number of strata was 30.
In view of the existing significant variation in the socio-economic characteristics in large cities in particular and in urban in general, each stratum of the large cities and urban strata was divided into four sub-stratum according to the socio- economic characteristics provided by the population and housing census with the purpose of providing homogeneous strata.
The frame excludes the population living in remote areas (most of whom are nomads), In addition to that, the frame does not include collective dwellings, such as hotels, hospitals, work camps, prisons and alike.
Sample Design
==========
The sample of this survey was designed, using the two-stage cluster stratified sampling method. The main sample was designed in 2009 based on the data of the population and housing census 2004 for carrying out household surveys. The sample representative on the Kingdom, rural, urban, regions and governorates levels. The total sample size for each round was 1336 PSUs (clusters). These units were distributed to governorates urban, rural and large cities in each governorate according to the weight of persons and households and according to the variance within each stratum. Slight modifications regarding the number of these units were made to cope with the multiple of 8, the number of clusters for four rounds was 53432.
The main sample is consisted of 40 replicates, each replicate is consisted of 167 Primary Sampling Units (PSUs). For the purpose of each round, eight replicates of the main sample were used. The Primary Sampling Units (PSUs) were ordered within each stratum according to geographic characteristics and then according to socio-economic characteristics in order to ensure good spread of the sample. Then, the sample was selected on two stages, in the first stage, The Primary Sampling Units (PSUs) were selected, using the Probability Proportionate to Size with systematic selection procedure. The number of households, in each primary sampling unit (cluster) served as its weight or size. In the second stage, the blocks of the primary sampling units (cluster) which were selected in the first stage have been updated. Then a constant number of households (10 households) was selected, using the random systematic sampling method as final PSUs from each PSU (cluster).
Sampling notes
==========
It is noteworthy that the sample of the present survey does not represent the non-Jordanian population, due to the fact that it is based on households living in conventional dwellings. In other words, it does not cover the collective households living in collective dwellings. Therefore, the non-Jordanian households covered in the present survey are either private households or collective households living in conventional dwellings. In Jordan, it is well known that a large number of non-Jordanian workers live as groups and spend most of their time at workplaces. Hence, it is more unlikely to find them at their residences during daytime (i.e. the time when the data of the survey is collected). Furthermore, most of them live in their workplaces, such as: workshops, sales stores, guard places, or under construction building's sites. Such places are not classified as occupied dwellings for household sampling purposes. Due to all of the above, the coverage of such population would not be complete in household surveys.
Response rates
The results of the fieldwork indicated that all sample households were visited. The number of successfully completed interviews was 48436, that is 90.8 percent of the total sample households.
Among the reasons of un-successful interviews (although three callbacks were made) 1.8 percent of the dwellings were closed at time of the visit.
The findings also indicate that the response rate is 95.5 percent, based on dividing the number of completed questionnaires by the number of expected completed interviews, that is after excluding the vacant dwellings.
More information on the distribution of interviews by region, governorate and visit results is available in table (E) in Page 4 of the annual report provided among the disseminated survey materials under a file named "Jordan 2014- Annual report (English).pdf".