Lecture 1 2 2 PDF

Preview

Summary

Lecture 1 Notes ...

Description

Section 4.1 Experimental and Observational Studies Anecdotal evidence - informal observations: different effect on everyone In an experimental study, a researcher conducts an experiment, by assigning subjects to certain experimental conditions and then observing outcomes on the response variable (or variables). The experimental conditions, which correspond to assigned values of the explanatory variable, are called treatments. In an observational study, the researcher observes values of the response variable and explanatory variables for the sampled subjects, without anything being done to the subjects (such as imposing a treatment). Lurking (confounding) variables are variables that drive both the explanatory and response variables It’s not possible to establish cause and effect definitively with observational studies. Association does not imply causation. Informal observations are called anecdotal evidence. A sample survey selects a sample of subjects from a population and collects data from them. It is a type of nonexperimental study. A census attempts to count the number of (all the) subjects in the population and to measure certain characteristics about them.

Section 4.2 Good and Poor Ways to SampleGo Sampling frame - list of subjects in which the population is taken from Sampling design - a specific method for selecting from the sampling frame Simple random sampling - selects n subjects from the population that provide equal chance of being selected Collecting an appropriate sampling frame must be executed in a certain area in which fits the questionnaire. Eg: Library: Ask the students how much they study? Faults with Sampling: - Bias: when the results from the sample are not representative of the population - Non-response bias: When subjects refuse or fail to complete all questions - Response bias: when the subjects gives an incorrect answer Eg: lying - Sample Undercoverage: having a sampling frame that lacks representation.

Section 4.3 Good and Poor Ways to Experiment Good -

-

Experimental units: The subjects of an experiment; the entities that we measure in an experiment. Treatment: A specific experimental condition imposed on the subjects of the study; the treatments correspond to assigned values of the explanatory variable. Explanatory variable: Defines the groups to be compared with respect to values on the response variable. Response variable: The outcome measured on the subjects to reveal the effect of the treatment(s). An experiment deliberately imposes treatments on the experimental units in order to observe their responses.

Randomization in experimental design: - To focus on the goal of the experiment to analyze the association between the treatment and the response for the population, not just the sample. - Use randomization for assigning subjects to the treatments - To eliminate bias that may result if you (the researchers) assign the subjects. - To balance the groups on variables that you know affect the response. - To balance the groups on lurking variables that may be unknown to you.

Control - A control group. A control group allows us to analyze the effectiveness of the primary treatment. - Placebo is also used as a control and may affect the results of the experiment Blind study - To measure and accurately achieve results to keep the subject blind from the study and minimize or eliminate the placebo effect which may produce bias results. - When neither the subject nor those having contact with the subject know the treatment assignment, the study is called double-blind.

Section 4.4 Other Ways to Conduct Studies (Partial) A case-control study is a retrospective observational study in which subjects who have a response outcome of interest (the cases) and subjects who have the other response outcome (the controls) are compared on an explanatory variable. This type of study is popular is medical studies however is not ethical. We cannot randomize subjects to a smoking group and non-smoking group as it would ask subjects to start smoking, thus cannot draw valid conclusions about general population. Prospective study identifies a group (cohort) of people and observes them in the future. Matched Pairs Design when we have two observations for a particular subject because they both come from the same person (sometimes called a repeated measures design). - A matched-pairs design in which subjects cross over during the experiment from using one treatment to using another treatment is called a crossover design. - Crossover design helps remove certain sources of potential bias. Using the same subjects for each treatment keeps potential lurking variables from affecting the results.

Section 2.1/2.2 Summaries of Data Frequency Table - Categorical Variable Proportion = frequency of that class/ sum of all frequencies Percentage = proportion x 100% Proportions and percentages are also called relative frequencies Categorical Data (Visual) - Pie Graph - Bar Graph Graphs for Quantitative Variables - Dot Plot - Frequency Histogram - Skewed to Left, Skewed to Right, Symmetrical...