Indonesia Survey Information 2009

Data Source
Estimates of media use in Indonesia are based on a sample survey of 3000 Indonesian adults (15 and older) interviewed face to face between March 6 and April 2 2009.

The target population was represented by the sample 143.438,000 adults. Interviewers made a total of 11,944 contact attempts in order to achieve the 4,000 interviews. The response rate for the survey was 63% as a proportion of valid contacts. The margin of error for the sample is ±1.5% with a 95% confidence interval. Percentages for subgroups carry larger margins of error, depending on the number in the group.

For sampling and weighting purposes, detailed information (gender, age, urban/ rural) is needed on the provincial level which is not always available for the most recent years. For this reason statistics from the 2005 Intercensal Population Survey are used describing demographics for the total population of 217 million. Estimation on Indonesia population in 2009 is 231 million based on projection being done by the Central Statistic Bureau (BPS) in year 2005.The sample was carried out in 21 provinces which account for about 93 percent of the population of the country.The main sample was divided into urban and rural in each province. The proportion of urban and rural is obtained from the urban and rural proportion of particular provinces based on 2005 Intercensal data. Each province was allocated X number of urban respondents and Y number of rural respondents so the final proportion was 43:57, matched with the 2005 Intercensal data. The discrepancy is corrected by applying weighting factors for the data.

Survey Content
The questionnaire covered a wide range of mass-media exposure and program interests, including the Internet and mobile-phone use.

Sampling
 The first step was choosing the provinces and districts. Proportionate to Population Size (PPS) method was used in this process. The Primary Sampling Unit (PSU) used was the Kecamatan (sub-district).  There are approximately 5,000 of these in Indonesia. In each PSU, 20 interviews were conducted. Therefore, for the main sample of n = 3,000, 150 Kecamatan (PSU) were selected. Once the list of kecamatan wasobtained, the next step was selecting desa (rural)/ kelurahan (urban) within each selected kecamatan. This process was based on list of desa/kelurahan from the General Population Census, 2000. The third step involves the selection of one RT (neighbor cluster) within each selected desa/kelurahan in each PSU. For the main sample, this process was conducted in the field, by the Field Supervisors. Household random selection was the same for main and booster sample. Once the RT (neighbor cluster) is selected, the interviewer approached the head of the RT to ask for permission to interview there. Once this is granted, the interviewer compiled a list of households within the RT. The starting point was selected using random number table also, so that every household had a chance of being selected (including Head of RT’s house).

The interviewer worked in a clockwise direction from the selected house to obtain the required five interviews, leaving an interval of three houses between contacts in rural areas and four houses between contacts in urban areas. If no contact/ house is locked/ refused, etc., the interval rule still applies.
    
Five interviews were allocated for each RT. If the number of interviews conducted in a selected RT was not sufficient due to many refusals or not enough households, the interviewer continued to the next nearest RT. Interviewers are required to draw a map of each RT, indicating the head of the RT’s house and the subsequent houses.  The map shows each subsequent household that is contacted, including those who refused to be interviewed and those that are empty.

Respondents within households were selected randomly by use of a Kish-grid method and a listing of adults in the household.

Indonesia Survey Information 2007/2008

Data Source
Estimates of media use in Indonesia were based on a sample survey of 3012 Indonesian adults (15 and older) interviewed face to face between December 15, 2007 and February 3, 2008.
For sampling and weighting purposes, detailed information (gender, age, urban/ rural) was needed on the provincial level which is not always available for the most recent years. For this reason statistics from the 2005 Intercensal Population Survey are used describing demographics for the total population of 217 million. Population growth is about 1.3% p.a. and the estimate for 2007 was a population of 223 Million. The sample was carried out in 21 provinces which account for about 93 percent of the population of the country. The main sample was divided into urban and rural in each province. The proportion of urban and rural wasobtained from the urban and rural proportion of particular provinces based on 2005 Intercensal data. Each province was allocated X number of urban respondents and Y number of rural respondents so the final proportion was 43:57, matched with the 2005 Intercensal data. The discrepancy was corrected by applying weighting factors for the data.

Survey Content
The questionnaire covered a wide range of mass-media exposure and program interests, including the Internet and mobile-phone use.

Sampling
 The first step was choosing the provinces and districts. Proportionate to Population Size (PPS) method was used in this process. The Primary Sampling Unit (PSU) used was the Kecamatan (sub-district).  There are approximately 5,000 of these in Indonesia. In each PSU, 20 interviews were conducted. Therefore, for the main sample of n = 3,000, 150 Kecamatan (PSU) were selected. Once the list of kecamatan wasobtained, the next step was selecting desa (rural)/ kelurahan (urban) within each selected kecamatan. This process was based on list of desa/kelurahan from the General Population Census, 2000. The third step involves the selection of one RT (neighbor cluster) within each selected desa/kelurahan in each PSU. For the main sample, this process was conducted in the field, by the Field Supervisors. Household random selection was the same for main and booster sample. Once the RT (neighbor cluster) is selected, the interviewer approached the head of the RT to ask for permission to interview there. Once this is granted, the interviewer compiled a list of households within the RT. The starting point was selected using random number table also, so that every household had a chance of being selected (including Head of RT’s house).

The interviewer worked in a clockwise direction from the selected house to obtain the required five interviews, leaving an interval of three houses between contacts in rural areas and four houses between contacts in urban areas. If no contact/ house is locked/ refused, etc., the interval rule still applies.
    
Five interviews were allocated for each RT. If the number of interviews conducted in a selected RT was not sufficient due to many refusals or not enough households, the interviewer continued to the next nearest RT. Interviewers are required to draw a map of each RT, indicating the head of the RT’s house and the subsequent houses.  The map shows each subsequent household that is contacted, including those who refused to be interviewed and those that are empty.

Respondents within households were selected randomly by use of a Kish-grid method and a listing of adults in the household.

Indonesia Survey Information 2006

Data Source
Estimates of media use in Indonesia are based on a sample survey of 3013 Indonesian adults (15 and older) interviewed face to face between November 10 2006 and December 12, 2006.
The Indonesian population of 214 million is spread over more than 14,000 islands. However, population density is not equally spread. About 80% of the population live on Java (59%) and Sumatra (21%) islands.
For sampling and weighting purposes, detailed information (gender, age, urban/ rural) was needed on the provincial level which has not always been available for recent years. For this reason statistics from the 2000 Census are used, describing demographics for the total population of 201.2 million.
Population growth was about 1.3% p.a. and the estimate for 2006 is a population of 217.5 Million. The sample was carried out in 21 provinces which account for about 93 percent of the population of the country.

The main sample was divided into urban and rural in each province. The proportion of urban and rural is obtained from the urban and rural proportion of particular provinces based on 2005 Intercensal data. Each province will be allocated X number of urban respondents and Y number of rural respondents so the final proportion is 43:57, matched with the 2005 Intercensal data. The discrepancy is corrected by applying weighting factors for the data.

Survey Content
The questionnaire covered a wide range of mass-media exposure and program interests, including the Internet and mobile-phone use.

Sampling
The first step was choosing the provinces and districts. Proportionate to Population Size (PPS) method was used in this process. The Primary Sampling Unit (PSU) used was the Kecamatan (sub-district).  There are approximately 5,000 of these in Indonesia. In each PSU, 20 interviews were conducted. Therefore, for the main sample of n = 3,000, 150 Kecamatan (PSU) were selected. Once the list of kecamatan wasobtained, the next step was selecting desa (rural)/ kelurahan (urban) within each selected kecamatan. This process was based on list of desa/kelurahan from the General Population Census, 2000. The third step involves the selection of one RT (neighbor cluster) within each selected desa/kelurahan in each PSU. For the main sample, this process was conducted in the field, by the Field Supervisors. Household random selection was the same for main and booster sample. Once the RT (neighbor cluster) is selected, the interviewer approached the head of the RT to ask for permission to interview there. Once this is granted, the interviewer compiled a list of households within the RT. The starting point was selected using random number table also, so that every household had a chance of being selected (including Head of RT’s house).

The interviewer worked in a clockwise direction from the selected house to obtain the required five interviews, leaving an interval of three houses between contacts in rural areas and four houses between contacts in urban areas. If no contact/ house is locked/ refused, etc., the interval rule still applies.
    
Five interviews were allocated for each RT. If the number of interviews conducted in a selected RT was not sufficient due to many refusals or not enough households, the interviewer continued to the next nearest RT. Interviewers are required to draw a map of each RT, indicating the head of the RT’s house and the subsequent houses.  The map shows each subsequent household that is contacted, including those who refused to be interviewed and those that are empty.

Respondents within households were selected randomly by use of a Kish-grid method and a listing of adults in the household.