Collecting Data Backbone of Statistics
It’s all about the Vocabulary! Population: the entire group that we are interested in Sample: some members of the group that are selected to represent the entire group Parameter of Interest: the thing we are interested in knowing about the population
Random Samples Simple random sample (SRS): each individual has an equal chance of being selected Ex: Numbered class list, number generator chooses 5 numbers Systematic sample: individuals are selected by a random starting point and then at regular intervals Ex: 50 people out of Start with person 100 and select every 20 th person Cluster sample: divide population into groups (clusters) then use all individuals in that group Ex: Sample from Treasure Valley to ask about cable rates. Divide into groups based on neighborhoods. Then select 10 neighborhoods at random and every house in the neighborhood is part of the sample Stratified random sample: divide individuals into similar groups that affect the response then take a SRS in each group Ex: Sample to see how many cookies they can eat in a minute. Separate the population into age groups then take an SRS of each group
Non-Random Samples Convenience sample: choosing easiest way to reach individuals Ex: Want a sample from Mountain View, so you interview the students in your 2B class Volunteer sample: individuals choose whether or not to respond Ex: Any internet survey
Which is Which? You are in charge of school activities. You want to know what activities students would prefer to participate in during the school year. You decide to put the name of each student in the school into a big bowl. You draw 100 names and ask those students to respond to a survey about the activities they prefer. SRS You are in charge of school activities. You want to know what activities students would prefer to participate in during the school year. You use the rolls from each homeroom class. You put the all the names from one class into the bowl and draw two names from the class. You go through each homeroom class, drawing 2 names from each class. You ask those students to respond to a survey about the activities they prefer. Stratified Random Sample You are in charge of school activities. You want to know what activities students would prefer to participate in during the school year. You get the list of all the homeroom classes and randomly select 5 classes. You go to each of the classes selected and survey all the students in that class. Cluster
Methods for Investigating Parameters Survey: a question or a series of questions asked to a sample to gather responses Observational Study: observes a sample and measures parameters of interest without changing anything Experiment: deliberately imposes some change or variable to measure responses Control group: group that does not receive any change Experimental group: group that receives the change
Which is Which? To determine whether drinking orange juice prevents colds, researchers randomly assigned participants to a group that drank no orange juice or a group that drank two glasses of orange juice a day. They measured the number of colds that each group had over the course of the year and compared the results of the two groups. Experiment Control: no OJ Experiment: two glasses of OJ To determine the effectiveness of a new advertising campaign, a restaurant asked every tenth customer if they had seen the advertisement, and if it had influenced their decision to visit the restaurant. Survey To determine whether exercise reduces the number of headaches, researchers randomly selected a group of participants and recorded the number of hours each participant exercised and the number of headaches each participant experienced. Observation