In my opinion, using statistical sampling in the US census would be better than using the census method. The following are the advantages of statistical sampling over census study: Statistical sampling is less costly compared to census. The number of units studied in any research project directly affects data collection and analysis costs. In most census studies, the number of units is prohibitively large than that of statistical samples. Secondly, the research takes longer to complete if it is a census study rather than sampling study.
Statistical sampling data is therefore, timely for decision making than census. Data that is obtained on untimely basis is not relevant for decision making, and use of such data may result to erroneous decisions. In most cases, well-calculated sampling study is likely to have a less total error than a census research. More detailed information is obtained from a sample population than from a census because statistical sampling takes less costly, takes less time and allows researchers to take extra care in the data processing stage(Partsons 278).
There would be bias if there was a live game, and the MVP survey was running. This is intentional bias. If somebody asking for the best football players when two teams are playing and the game being aired live, the broadcasters would be causing a possible bias in that response. This is because of the of the following reasons: First, it is possible that the broadcaster will be asking a lot of people who are fans of either of the teams playing since they are the most interested in watching the TV at that particular time, and not all the viewers. This would undoubtedly result to the majority of them picking a player from that particular match.
Secondly, even if a person is not a fan of football or even is not familiar with the league, the presence of the match would influence him to vote for the outstanding player at that match. The bias would be eliminated by running the survey when no live match is being broadcasted or at the end of the league since all teams will have played and thus the audience will have known the outstanding player from all the games. Secondly, the broadcaster can plan, for a session, to highlight all the games played in the league and then ask the audience to vote.
Surveys, especially self-administered, surveys are relatively inexpensive. It is possible to administer surveys and get feedback from remote locations using email, mail or telephone, thus making it a relatively undemanding and flexible method of acquiring information. Consequently, the researcher can be able to reach out to a large sample, which makes the results statistically significant even with analyzing multiple variables. Additionally, it is possible to ask the respondents many questions about a given topic, and this gives considerable flexibility to the analysis. Since conducting surveys involves the use of similar written questions to all respondents, the standardized questions make the analysis more precise since uniform definitions are enforced upon the participants. Usually, high reliability is easily obtained due to standardization.
The weakness of survey method is that it is hard for participants to give correct feedback about a controversial question, the researcher is likely to get minimal reply from the selected sample and it forces the researcher to ask general questions, which may not be applicable to all respondents. A loaded question is a question where the researcher seems to expect a predictable outcome, other than just “yes” or “no” answer for example: “did you enjoy spoiling the breakfast for everyone else?” It is appropriate to use focus groups in situations where the desired information about motivations and behaviours is complex than the questionnaires are likely to reveal. For instance, in case of a situation when there is organizational conflict, members of then focus group may reveal in-depth information since they feel listened to.