Data Analysis Quiz 2

Quiz 2 Question: When faced with a data analysis problem that involves two variables, explain how you would decide which graphical display and numerical measures to use. Here are your answers in random order. See my comments. identify explanitory variable and response variable, and categorical and quantative. Good. The graphical display will depend on the kind of variables. For example, if the problem deals with two quantitative variables, then a scatterplot will be used. If there are two categorical variables a two-way table will be used. Also if one variable is categorical and the other quantitative, side-by side boxplots will be used. Good. I would decide which graphical display and numerical measures to use by determining which case the problem applies to. Good. I would decide which type of graphical display and numerical measures by deciding if the variables are categorical or quantitative. Good. You look at the explanatory and response then which is categorical or quantative and ddecide based on that. Good. Use a side-by-side box plot to graph the data and use the histogram to see it better with the fivenumber-summary... No, this is not correct. < No Response Given > < No Response Given > I would decide which graphical displays and numerical measures to use depending on the case number. For example, if I had case III, a quantitative explanatory and a quantitative response, I would use scatterplots and correlation coefficients, etc... to determine the problem Good. I would first of all classify its case. Ex Case 1,2,3,4. Then from that I would conclude what type of visual display to choose from. For numerical measures I would choose either percentages for categorical, and numerical values or real numbers for scatterplots. Your last sentence doesn’t make sense, but the first two are correct. < No Response Given > First we have to find which case the variavble belongs to, and afterwards it would be easer to find the graphical display. Good. side-by-side plots(relationship between categorical explanatory and quantative response), two way table (relationship between categorical response and categorical explanatory), scatterplot(relationship between two quatative variables) This is good, but doesn’t really answer the question. But I think you got the idea. If case I, where the explanatory variable is categorical, and the response variable is quantitative, then I would use side-by-side boxplots. If case II, where both variables are categorical, a twoway table works best, but if working with the conditional percentages, the information could be put into a double bar chart. If case III, where both variables are quantitative, a scatter plot works best. Good. It depends on the type of variables are presented. Once it is established that the variables are either categorical or quantitative then an appropriate graphical display can be chosen. Good. define the variables (explanatory or response) and also define they are categorical or quantitative variable. Good. To decided which graphical display and numerical measures to use, you first have to figure out which case you are dealing with: either I, II, or III. When you figure that out, you can then decide which graphical display and numerical measure to use. Good. Firstly, you would have to decide if the each variable is either a categorical response & a categorical explanatory (case ii), quantitative response & categorical explanatory (case i ), categorical response & quantitative explanatory (case iv) or quantitative response and quantitative explanatory (case iii). Good. first you have to figure out which variable is quantative or categorical and also which is explanatory and response. then figure out which case it falls under there you will figure out which graphical display and numerical measures to use. Good. i will use the box plot because it represents the percentage of the data No. This is not correct and doesn’t answer the question. < No Response Given > The relationship between a categorical and quantitative variable is Case I which uses side by side box plots with the 5-number summary. The relationship between two categorical variables is Case II which uses a two way table with conditional percentages. The relationship between two quantitative variables is Case III which uses a scatter plot with direction, form, and strength (correlation coefficient and least squares regression line). Good, but doesn’t really answer the question. But I think you got the idea. Well graphical i would use the two way table. To see the percentages. I would also use numerical measures to see how it comes out in numbers. Would you use tow-way tables ALL the time? And the last part of your answer doesn’t really make sense. Depending on the types of variables (explanatory vs. Response) determines which graphical display and numerical measures are used. Good. Once you figure out which variable is the explanatory one and which one is the response variable and whether it is categorical or quantitative, you can use the chart the role-type classification table to help you. For example, when it comes to a categorical explanatory variable and a quantitative response variable, it would be a Case I problem in turn letting us know that the best way to show this data graphically would be through the use of side-by-side boxplots using desrcriptive statistics as our numerical measure. It is the same process for the rest. Very good. The best way would be to determine which case the variables fall into, 1,2 or 3. If its two quantitative variables, then the obvious choice will be a scatterplot. Good. figure out whether the response and explanatory variables (data) are quantitative or catagorical, which "case" they fit into and then proceed appropriately. Good. i can use either scareplot This doesn’t make sense. I would use the Role-Type Classification table when faced with a data analysis problem that involves two variables. I would then determine which case it falls under and work on the problem from there on. Good. based on what type of explanatory and response variable your given you would pick between the graphs Good. It depends whether the explanatory and response variables are categorical or quantitative. Case 1- five number summaries, means and sd's. Plus side by side boxplots. Case 2- two-way tables, conditional column or row percentages. Side by side bar charts. Case 3- correlation coefficient, LSRL. scatterplot. Good. I would first figure out which variables are explanatory variables and which are response variables. After that,I would find out if the variables are categorical or quantitative. I would then apply each one to the appropriate case which would tell me the correct graphical display and numerical measures to use. Very good. < No Response Given > It depends on which two variables you have and then accordingly you will choose the graphical display and numerical measures. Good. You would use one of the various graphs depending on the variables. Categorical and explanatory variables determine the case which determine the display that should be used. Good. I would first figure out the explanatory and response variables and whether they were categorical or quantitative, then I would see which case it was (I II or II) and choose the display accordingly. Very good. When exploring the relationship between a categorical explanatory variable and a quantitative response variable (case 1) you use side-by-side boxplots and the numerical measures is the 5 number summary. When exploring the relationship between two categorical variables(case 2) you use two way tables and the numerical measures are conditional percentages. When exploring the relationships between two quantitative variables (case 3)you use a scatterplot and the numerical measure is the correlation coefficient. Good. First define if the variables are categorical or quantitative. Then depending on if the data is approximately symmetrical, you decide which measure to use. Your first sentence is correct, but the second one doesn’t make sense for two variables. First we must determine the explanatory variable and the response variable. Then we must decide which case they belong in using the Role/Type Classification. After determining which case they are, then we can decide which graphical display and numerical measures we can use. For example, if it was Case II, categorical and categorical variable, then we can use side-by-side bar charts or two-way tables. Very good.