Cor578.asc 5/18/95 Updated 08/30/02 Processing of Transfers to Children Questions from INTR.Q (Questions 660 through 701b) The questionaire contains seven different reasons that respondent may have transferred money to a child. For each reason, the series begins by asking if any children were given money for that reason. If the answer is yes, then which child received the most, the next most, etc. is obtained. If the answer is no, then the questions continue with the next reason. For any child who received money, the total amount received and the most recent year are also determined. The survey instrument allowed for transfers to children for each of these reasons: down payment on a home (3), enter or continue business or farm (3), increase wealth or decrease debt (3), medical expenses (4), housing or other living expenses (4), educational expenses (3), and other (4). The number in parenthesis indicates maximum number of transfers allowed for that reason. The total number of transfers in the actual data was 16. For some reasons data was defnitely missed due to the ceiling on number of transfers. Of those graduates who have 4 or more children: 22.74% reported transfers to 3 children for educational purposes; 4.22% reported transfers to 3 children for increasing wealth or reducing debt; 2.64% reported transfers to 3 children for a down payment on a home; and .33 % reported transfers to children to start a business. Of those respondents who have 5 or more children: 1.09% reported 4 transfers for "other"; .7% reported transfers for living expenses; .23% reported transfers for medical expenses. Since we have a children's roster with other characteristics of each child, we chose to match the child who received money with the child in the children's roster to learn more about these children. A computer program matched 90% of these names. Another 7% were matched by hand coding. For 1% of the remaining match attempts, there was a name given as a response for the transfer, but no match could be made with the children's roster. The reason no match could be made were categorized as follows: a. Name given, but not listed in children's roster b. More than one child listed in response to same question. Unclear if amount is total for all children or per child. c. Ambiguous response such as son, when there is more than one son in the children's roster. (If the response was son, when there was only one male child in the children's roster, that child is coded as the recipient. And a response of daughter was handled similarly.) d. Step child, not listed when the children's roster was collected. A separate code was assigned for each of these reasons. The remaining 1% of names were refusals. (Due to rounding total % is 99, not 100). If the name could be matched with the children's roster, then the variable RQ027 was assigned the position of that child in the children's roster. See COR524 for complete details concerning the children's roster. Basically, the children's roster can have up to 14 children per respondent. The set of variables are arranged from oldest to youngest. So if RQ027 is 3 for example, then the transfer was to the third oldest child and that child's characteristics can be found in the third set of variables from the children's roster. The sex and whether or not the child was the selected child were carried along with other responses in this section, for ease of analysis. It seemed more useable to have one table with sufficient space for all transfers to children and the reason for transfer included as a variable. The maximum number of transfers to children was 16. So a table with 16 entries of the following variables was created: RQ027 - birth order of this child RQ028 - is this child selected child RQ029 - gender of this child RQ030 - amount respondent gave to child RQ031 - why respondent gave money to child RQ032 - year respondent gave money to child After the table was created it was sorted by RQ030, amount respondent gave to child. Missing values for amount were put at the end of the table. If the respondent replied with don't know or refused to the initial question about transfer to children for the first reason, then there is a skip to the second reason, and if respondent replied with don't know or refused to initial question about transfer for the second reason there is a skip to the third reason, etc. If the respondent replies with refusal to all these initial questions about transfers to children, then all the above variables are coded refused. There are 246 of these cases. similarly if the respondent replies don't know to all initial questions about transfers to children, then the code is don't know for all variables. There are 4 of these cases.