Www.birdvilleschools.net



AP Statistics Project: Response Bias in SurveysThe Project: You and your partner (optional) will design and conduct an experiment to investigate the effects of response bias in surveys. You may choose the topic for your surveys, but you must design your experiment so that it can answer at least one of the following questions: Can the wording of a question create response bias? Does providing additional information create response bias? Do the characteristics of the interviewer create response bias? Does anonymity change the responses to sensitive questions? Does manipulating the answer choices/order of answer choices change the response? Can revealing other peoples’ answers to a question create response bias? You should compare at least 2 different groups. Usually this is a “control” question that is “normal” and unbiased and then the second group that is biased in a certain direction. Depending on your idea, you may choose to have 2 different biases—one that will tilt in one direction and one that will tilt in another direction. You will not be penalized if you do not succeed in creating a large bias. Due Dates: Proposal: Friday, October 27th Extra Credit early deadline: Friday, December 1st Report, Visual Aid, & Presentations: TBD—right before semester exams Proposal (20 points): The proposal is worth 20% of the grade, so don’t take it lightly. If the proposal isn’t approved the first time, you will need to resubmit it for a reduced grade. You must attach the original proposal to any resubmissions. In your proposal, you should include: A definition of the populationA copy of your survey questionsA short description of how you will create bias & what direction you think the bias will go (hypothesis)Where and how you will collect your dataNOTE: Your sampling procedure should NOT be biased. Survey 40 people per question (40 unbiased, 40 for each bias) If your idea is difficult to survey, I may reduce your required sample size. Visual Aid (30 points): The purpose of the visual aid is to support your presentation. It should be pleasing to the eye. The colors on your graphs are crucial to communicating your bias: use a consistent color-key so the change can be easily spotted. ALL electronic visual aids must be on my computer and given a practice run by Wednesday, December 7th. Oral Presentation (10 points): Each pair (or individual) will be required to give a ~5 minute oral presentation to the class. Both members need to participate equally and should be prepared to answer questions. Report (40 points): Typed, professional, about 1-2 pages. The report should include the following: Title (in form of a question)Introduction: In the introduction, you should discuss what question(s) about response bias you are trying to answer, why you chose this topic, and clearly state your hypothesis. Data Collection: In this section, you will describe how you obtained your data. Be specific. Graphs and Summary Statistics: Make sure the graphs are well labeled, easy to compare, and help answer the question of interest. The graphs should “stand alone”!Discussion and Conclusions: In this section, you will state your conclusions. You should also discuss any errors you made, what you could do to improve the study next time, and any other comments based on your critical reflection on the project. SEE BACK FOR RUBRICResponse Bias Project Rubric4 = Complete3 = Substantial2 = Developing1 = MinimalProposal (20%)Clearly defines the population being surveyedWell thought out survey questionsClearly states procedures for collecting dataReasonable hypothesis of the expected response biasStates populationSurvey questions are includedHypothesis includedStates populationSurvey questions OR hypothesis includedBriefly describes context but missing detailsVisual Aid(30%)Well organized, neat, and easy to readIncludes appropriate graphs/tables/etc. with consistent color-key for easier comparison of dataVisually appealing & creativeOrganized, neat, & easy to readIncludes appropriate graphs, tables, etc. Unorganized and/or hard to read Includes graph and/or tableVisual aid has severe problemsIntroDescribes context of researchClearly states questionProvides hypothesisQuestion of interest is of appropriate difficultyDescribes context of research with specific question of interestSuggests hypothesis OR has appropriate difficultyIntroduces context of research of interestHas specific question OR has question of interest and hypothesisBriefly describes the context of the researchData CollectionMethod is clearly describedIncludes appropriate randomizationDescribes efforts to reduce bias, variability, confoundingQuantity of data is appropriateMethod is clearly described Some effort to incorporate good data collectionQuantity is appropriateMethod of data collection is describedSome effort made to incorporate good data collectionSome evidence of data collectionGraphs & Summary StatisticsRaw data is included in two-way table (categorical) or in two lists (quantitative)Appropriate graphs included, graphs are neat, clearly labeled, & easy to compareAppropriate summary statistics are included in discussion (e.g. percentages for categorical data, means for quantitative data, etc.) Appropriate graphs are includedGraphs are neat, clearly labeled, easy to compareAppropriate summary statistics or raw data includedGraphs and summary statistics are includedGraphs OR summary statistics are includedConclusionUses results of study to correctly answer questionDiscusses what inferences are appropriate based on study designShows good evidence of critical reflection (discusses possible errors, shortcomings, alt. explanations, etc.)Makes correct conclusionDiscusses what inferences are appropriate OR shows good evidence of critical reflectionMakes a partially correct conclusionShows some evidence of critical reflectionMakes a conclusionPresentation(10%)Has a clear, holistic understanding of the projectSpeaks clearly & with confidenceDoes something to engage the audienceHas clear understanding of projectSpeaks clearly & with confidenceDoes not have full understanding of projectCommunication is unclear/unorganizedCommunication is unorganized & poor -372583-2838450Report (40%)00Report (40%) ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download