Testing Variables

international organizations are often to be considered as risky, because donors also might take extra considerations in their decision of donating to countries with conflicts, lesser openness and/or lesser democracy. As indicated in paragraph 2.2.2, the CBF organizations nowadays rates efficiency, which is a threshold of 70 % over a period of three years, but does not rate project performances.

3.2 Testing Variables

We justify the way of how the conversion was calculated from the fact that respondents may be a little enthusiastic and optimistic about their giving behaviour, while this is an experiment and people do not always act the way they say they do. Therefore, for most cases we took the average number and tried to be conservative regarding the larger amounts and numbers.

This approach stems from research by Pace (1949) where trained interviewers conducted surveys with questions involving elections and donating, whereas personal questions indicating prestige (owning a car, driver’s license, a house or a telephone) and desired social behaviour (voting and donating to elections) were involved.

The research checked the answers from respondents with known data from

governmental and local agencies and found discrepancies between the data from the agencies and the respondents’ answers, varying from nearly zero to half of the answers received. Many respondents exaggerated in their answers, specifically concerning the election questions.

We therefore divided the times of giving by a half, except for the last option where we chose to keep it at 12 times because it is more likely that people donate around 12 times a year, than many more times. If they wish to give that often it may be more logical to give higher amounts. The only situation where we can think of is when people visit an organization such as a church and give each time upon their visit.

When organizations are asking for funds, there are always psychologically favourite amounts like 7,5; 35 and 75 euros which found their origins in the Dutch money (coins and bills) that were widely used to the year 2001 for guilders: 1; 2,5; 5; 10; 25; 50; 100; 250; 1000 and which are still labelled in the heads of people. After 2001, the euro was introduced and used 2; 5; 10; 20; 50; 100; 200; 500. While nowadays, cash money is less frequently being used, we still expect people to calculate with these amounts.

Former intervals with amounts

Amount

Lesser than 5 euros (1) 2.5

5 to 10 euros (2) 7.5

10 to 25 euros (3) 15.0

25 to 50 euros (4) 35.0

50 to 100 euros (5) 75.0

100 to 250 euros (6) 175.0

250 euro or more (7) 250.0

Former intervals with n for giving: N giving

once (1) 0.5

once a year (2) 1.0

2 to 5 times a year (3) 3.0

6 to 12 times a year (4) 8.0

often than 12 times (5) 12.0

When people are shown the intervals with the amounts and are asked which they are most likely to donate to an organization, we expect them not to know exactly which amount they would really donate. It may be the lowest amount of the interval but could also be the amount of the upper level, so therefor we took the average of the lower and upper value of the interval and kept to a slightly lower amount in the cases of the 10 to 25 euros (15 euros), while the minimum of 250 euros was kept at 250 euros. In the cases where donors have chosen for giving more than 12 times to an organization and when the amount of 250 euros incurred, the conversions were kept to the lowest boundary of the interval.

Another indication of how to rate these amounts can be found when visiting a website of charity organizations and ask how to donate. On the website of Oxam Novib (Oxfamnovib, 2021) the following amounts are occurring in euros for monthly donations: 6; 10; 15; 20; 50 and 25, 50, 100, 250 500 for incidental donations. With the organization KWF(who fights against cancer), the monthly donations are in euros: 6; 10; 12 and amounts to be given incidentally: 15; 25; 50 (Kankerbestrijding, 2021).

3.2.4 ‘No treatment’ and ‘treatment’ groups

Initially, the four treatment groups were labelled as categorical independent variables because there is no ordering between the groups. However, for our research and hypothesis we

investigated whether or not reporting on project performance has an effect on donating behaviour from no donors, non-regular donors and regular donors while project performances could turn out to be positive, neutral and negative towards budget.

The categorical dummy variable ‘Treatment’ was made to distinct between the groups 1, 2 and 3 which were offered analytical graphs with project performance and group 4 which had not been offered the additional graph, the ‘no treatment’ group. The ‘no treatment’ group, serving as control group was labelled with value ‘zero’ and the groups that were offered the analytical graphs, received value ‘one’.

3.2.5 Demographic data

In our theory we discussed earlier research regarding donating behaviour which included objective human traits which could predict donating to charities. Demographic information such as age, gender, education and net income are objective independent variables which are included in our testing.

3.2.6 Other data

Donations that were done in the past by charity category (question 14) were converted in the same way as was done for new donations resulting from the (no) treatment questions 10, 11, 12 and 13. The totals for each category of charity were calculated by multiplying times of giving by the donated amount on a yearly basis, whereas the totals of each category were summed which resulted into the independent variable ‘TotalAmountFormerDonations’.

To support our calculations, we refer to paragraph 3.3.3 concerning the calculation of new donations, because the earlier mentioned questionnaire of Pace (1949) also involved questions regarding past behaviour, whereas one specific question considered voting on six different elections that went back up to three years before, indicating an accuracy of correct reporting of about 70 %, which might stem from exaggerating regarding preferred social behaviour but also from the fact that the human memory may be not as accurate as expected.

Koriat, Goldsmith and Pansky (2000) did a research for the accuracy of human memory and found several categories of “how memory can depart from what actually occurred: falsely recalling, wrongly combining, distorting remembered information,

remembering information at different levels and metamemory errors”, of which we address the fact that when respondents belong to a household with more individuals and do not perform the actual act of giving, they may think certain donations were done and under- or overestimate their households’ giving (p. 517).

On the other hand, as the survey was sent out around end of May it may be very likely people still remember their donations from the Dutch income tax form which was due this year before the 1^st of May. How much people remember from their tax form also depends on whether or not they were able to deduct the amounts that were given, for the Dutch tax system works with certain thresholds in order to control for incomes, whereas higher incomes require larger amounts from donations than lower income groups to be allowed to deduct these from income tax.

While remembering of what was formerly given is somehow controlled for, we are still dealing with the possibility of exaggerating donations by respondents and therefore, for the independent variable concerning total amount of old donations we will stick to the same table as we used for the dependent variable that concerned total amount of new donations.

3.2.7 No givers, non-regular givers and regular givers

An extra independent value ‘donortype’ was made to distinct between non-givers, non-regular and regular givers. Because we claim former non-regular givers and regular givers to be

different, as non-regular givers feel lesser of a bond (Anderson & Miller, 2003) with the organizations they are donating to, we thus separated the group of givers into non-regular and regular givers, for which we use the earlier calculated number of donations and amounts of former donations as an indicator to distinct between the groups.

People who gave more than four times a year are to be considered as regular givers while people who gave lesser ought to be considered as non-regular givers. Although there appears to be given seemingly ‘insignificant’ amounts, the amounts were kept in the tests, while it is hard to judge if a certain amount is insignificant, for it is a personal donor’s judgement if he/she can miss the amount from their household budget. From our point of view, the idea upon judging whether a donor is a regular giver stems from the fact that someone is deliberately willing to give on a regular basis and has developed a habit for it, which separates a non-regular giver from a regular giver.

In addition, people who were assigned by Prolific to be no donors to charity, reacted positively to our treatments by willing to donate. Therefore, we will consider this group as the (former) ‘no donor’ group.

Three categorical dummy variables were made and labelled as ‘no donor’, ‘non-regular donor’ and ‘‘non-regular donor’. Respondents who fell into either of the three categories, were assigned with value ‘1’ for that category and value ‘0’ if not.