Census Data

1277 Words6 Pages

1. I would analyse a combination of both census data and historical/archived data. This type of quantitative data will contain the socio-economic status of parents on how well their school-aged children do in school. Census data in Canada is readily available and very cost effective for me to use. The census does not contain direct information that links how well a child does in school to the socio-economic status of their parents, but there is other data the can be used to form the connection. Some school boards even post the average income levels for schools. Then it can be compared to the standardized testing results if the province does these test. Depending on the time period I am examining, I may be required to use historical/archived …show more content…

Taking a quota of sample respondents will provide more accurate data when studying the income gap between men and women in Canada. It will provide a more accurate sample to use since I am able to control certain variables. For example, using this method will allow me to ensure that different fields of work are represented proportional to their size. However, there is still the risk of sampling bias and the data I am using to get the quota from has to be up to date and a reliable source. I am still limited to the choice that people have to participate or not, and by budget constraints. Using a sample of easily available respondents will save the most amount of time and money. So they are great if budget is a concern. Data can be collected quickly since there is a smaller need to find and select people to be in the sample. However, when studying the income gap, this method will give the data sampling bias, more than the quota method. Variables such as occupation, and experience cannot be controlled, making it difficult to accurately analyze the …show more content…

Using a stratified random sample of respondents to study the relationship between voting behaviour and higher education will reduce the sampling error greatly, even to zero depending on what is known. Since a larger selection of respondents is selected, the amount of variables is limited, which decreases sampling bias. In this case, I can group individuals by their level of education, select a proportional sample, and learn their voting behaviour. A disadvantage to this method is that I am limited to the data that is available for me to use. If reliable and up to data is not available, then my findings would not be reliable. Budget and time constraints also are an issue due to the potentially large amount of respondents. Using a sample of easily available respondents will be the quickest method to gather data, saving time and money. However, this method has drawbacks. There is sampling bias since it depends on availability. This also makes it more difficult to find an accurate sample of respondents with variety. For example, if the easily available respondents are all hippies, then it will not produce good data to study. 5. Standard error is 0.07 or 7%. 52% women. Let x be the unknown
0.07 =

Open Document