Zogby International conducted interviews of 600 people in India (Hindu, Muslim), Peru (Roman Catholic), Russia (Russian Orthodox), Saudi Arabia (Muslim), and South Korea (Buddhist, Christian); 593 in Israel (Jewish, Muslim, Druze); and 795 in the United States (Catholic, Protestant). All interviews in India, Peru, Russia, Saudi Arabia and South Korea were conducted face-to-face. All interviews in Israel and the Unites States were conducted over the phone. All calls for the United States were made from Zogby International headquarters in Utica, N.Y. All interviews -- face-to-face and telephone -- were conducted from January through March of 2003. The margin of error for India, Peru, Russia, Saudi Arabia, South Korea, and Israel is +/-4%. The margin of error for the United States is +/-3.6%. Margins of error are higher in sub-groups.

The total sample size was specified to be 600 interviews with questionnaires to be filled out in the capital, New Delhi, with the two main localities (i.e. religions) Hindu (70%) and Muslims (30%). The gender composition was based on that of the Indian population (above 15 years of age), with 52% males and 48% females, applied to each locality. Therefore, one can say that a stratified random sample selection was adopted with probability proportional to the size (PPS) of the population in case of selection of males and females.

The sample selection was based on a list of the electoral polls of the Delhi region provided by the census of the Office of the Registrar General India. To select the required number of subjects from the list, a Systematic Random Sampling Scheme was adopted.

The total sample size was specified to be 600 adults from the general population interviewed nationwide by telephone. Telephone numbers were randomly selected. Israelis composed 90% of the sample, while Palestinians made up 10% of the sample. Interviews were conducted in Hebrew or Arabic. Calls were made from January 11-16, 2003.

The total sample size was specified to be 600 interviews with questionnaires to be filled out in the capital, Lima, interviewing only Catholic Christians.

Interviewers resorted to public places chosen from different social neighborhoods and workplaces, which may result in a "selection bias." The choice of interviewees in each coffee shop and workplace was random. As for the gender, it was controlled when choosing the interviewees to reflect the gender composition of Peru's population (above 15 years of age), with 50% males and 50% females.

It may be worth noting that there is a comparison limitation to this methodology because comparisons can only be applied to other countries' middle-class population, as it is a common behavior for middle-class people to go to coffee shops. Generally, the upper and lower social classes do not visit these public coffee shops.

The total sample size was specified to be 600 interviews with questionnaires to be filled out in the capital, Moscow, interviewing only Russian Orthodox.

Interviewers resorted to public places chosen from different social neighborhoods, such as coffee shops, which may result in a "selection bias." The choice of interviewees in each coffee shop was random, with not more than one person from each table. Gender was controlled when choosing the interviewees to reflect the gender composition of Russia's population (above 15 years of age), with 46% males and 54% females.

It may be worth noting there is a comparison limitation to this methodology because comparisons can only be applied to other countries' middle-class population, as it is a common behavior for middle-class people to go to coffee shops. Generally, the upper and lower social class people do not visit these public coffee shops.

The total sample size was specified to be 600 interviews with questionnaires to be filled out in both Riyadh and Jeddah, with a sample distributed as 60% and 40%, respectively, which was based on Information International's experience in Saudi Arabia and the population size of both cities. The main religion in Saudi Arabia is Muslim and, consequently, all the 600 interviewees are Muslims. The gender composition of the sample was based on that of the Saudi population (above 15 years of age) with 59% males and 41% females. Consequently, and due to the lack of a population census, the sample was chosen based on a Systematic Random Sampling Scheme within the three different strata of economic social class, whereby certain starting points inside each stratum were specified (i.e.36 starting points were chosen in Riyadh and 24 in Jeddah) and 10 random household interviews around each starting point were conducted.

Therefore, one can say that a Stratified Random Sample Selection was adopted with probability proportional to the size (PPS) of the population in case of selection of males and females.

The total sample size was specified to be 600 interviews with questionnaires to be filled out in the capital, Seoul, with the two main localities (i.e. religions) Christian (50%) and Buddhist (50%). The gender composition of the sample was based on that of the Korean population (above 15 years of age), with 49% males and 51% females. Therefore, one can say that a stratified random sample selection was adopted with probability proportional to the size (PPS) of the population in case of selection of males and females.

The sample was selected based on a Systematic Random Sampling Scheme after specifying certain starting points (i.e. areas) for both Christian and Buddhist areas that also take into consideration the three different strata of economic social class.

Zogby International conducted interviews of 795 adults chosen at random nationwide. All calls were made from Zogby International headquarters in Utica, N.Y., from February 19 to February 21, 2003. The margin of error is +/- 3.7%. Slight weights were applied to age, gender, region and religion to more accurately reflect the adult population. Margins of error are higher in sub-groups.