Please enable JavaScript.

Coggle requires JavaScript to display documents.

Producing Data (Observational Studies and Experiments (Vocabulary of…

- - - - Definition: A voluntary sample is a sample where people elect to participate. This creates a bias because the people who volunteer to participate in a survey usually have specific or strong opinions.
      - Example: People are asked to call in to the radio and whether they are pro-life, pro-choice, or neutral. There is a good change that anybody calling in have strong opinions, and are not neutral
    - - Definition: A convenience sample is made up of people who are easy to reach. This sample may create a bias in the data.
      - Example: A survey is conducted to determine the favorite foods of people in Duxbury, Massachusetts. The surveyor recruits the first 50 people who walk outside of a Chinese food restaurant.
    - - Definition: Undercoverage is a type of selection bias that happens when not all members of a population are represented in a sample.
      - Example: A survey is conducted to see if people in Duxbury would go to watch Sunday morning sports games. The surveryor stands outside of Holy Family and asks people leaving the church. This population would be heavily leaning towards not attending the games because most of them would want to be at church at that time.
  - - - Definition: a response bias is a bias that comes from issues in the data-taking process. This can include questions that are designed to produce a certain answer, or people chosing answers that they believe is correct, not what they actually think
      - Example: A survey asks people their thoughts on animal shelters that euthanize animals after a certain amount of time. The results may be skewed because participants don't want to appear mean or insensitive
    - - Definition: a non-response bias occurs when participants chose to not answer the survey. This can skew the results by only including answers from people willing to
      - Example: A random sample is selected to participate in a phone survey regarding their thoughts on racism. The people who chose to answer are more likely to have strong thoughts on the topic which will skew the results.
    - - Definition: Wording bias is a survey error in which the wording of the question may effect the answer. The words in the question have certain connotations that influence people differently.
      - Example: "Do you believe in the dealth penalty?" versus "Do you believe human beings should be murdered for their crimes?"
  - - - Definition: A SRS occurs when all members of a population are equally likely to be chosen for the study.
      - Example: If a survey was being conducted for 50 students from the DHS Class of 2018, each person could be assigned a number from approximately 001-320, and we could use a random number table to randomly select 50 students.
      - Advantages: SRS eliminates the risk of any sampling biases. It is impossible to misrepresent a population if the participants were chosen at random.
      - Disadvantages: SRS can limit the number of participants in a study, because it takes time and money to assign a number to each individual.
    - - Definition: A stratified SRS is a type of SRS in which the entire population is divided into groups called strata, and then a SRS is performed on each strata.
      - Example: If a study is done to determine the food preference of airport patrons, they can divide the tables in the food court into stratas, and randomly select participants from each strata to be studied.
      - Advantages: Stratas are advantageous for large populations where it is not feasible to study all individuals. The study is random so it still represents all individuals without needing to survey each one.
      - Disadvantages: a stratified SRS requires more time and effort for the researcher.
    - - Definition: In cluster sampling, the population is divided into representative groups, and certain groups are selected to participate in the study.
      - Example: When studying food preference of airport patrons, the researcher can divide the food court into groups , and pick certain groups for participation.
      - Advantages: Instead of performing SRS on individuals, the researcher only has to randomly select groups, which greatly reduces the amount of time.
      - Disadvantages: The researcher needs to ensure that each group is completely representative. In the food court example, if the clusters chosen are the one closest to Popeyes, and the one closest to Papa Ginos, there might be an accidental sampling bias.