STAT METH FOR RSRCH
STAT METH FOR RSRCH STAT 401
Popular in Course
Popular in Statistics
This 12 page Class Notes was uploaded by Giovani Ullrich PhD on Saturday September 26, 2015. The Class Notes belongs to STAT 401 at Iowa State University taught by Staff in Fall. Since its upload, it has received 25 views. For similar materials see /class/214407/stat-401-iowa-state-university in Statistics at Iowa State University.
Reviews for STAT METH FOR RSRCH
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/26/15
Correlations Year Yeaqur US Population data Correlations between explanatory variables Year 10000 09999 Yeaqur 09999 10000 ZUUU 1950 1900 1850 1800 Year 4000000 3900000 3800000 3700000 3600000 3500000 3400000 3300000 Yeaqur I I I I I I I I I I I I 1800 1850 1900 1950 2006300000 36000003800000 Pairwise Correlations Variable Yeaqur by Variable Year Correlation Count 09999 21 Signif Prob 00000 Correlations Year YearCtqur Year 10000 00000 YearCtqur 00000 10000 Scatterplot Matrix 1950 l 1900 Year I 1850 l 1800 39 10000 I l 8000 I I 6000 I YearCtqur 4000 2000 0 l 2000 I 39 I I I I I I 1950 2000 2000 4000 6000 8000 I 39 I 39 I 1 800 1 850 1 900 Highway MPG for Various Types of Drive Multiple Regression Analysis Response Highway MPG Explanatory All Wheel indicator variable Rear Wheel indicator variable Summary of Fit RSquare 0254103 RSquare Adj 0238723 Root Mean Square Error 5311626 Mean of Response 277 Observations or Sum Wgts 100 Analysis of Variance Source DF Sum of Squares Mean Square F Ratio Model 2 9323031 466152 165224 Error 97 27366969 28213 Prob gt F C Total 99 36690000 lt0001 Parameter Estimates Term Estimate Std Error t Ratio Probgtt Intercept 29983333 0685728 4372 lt0001 All Wheel 7374638 1302648 566 lt0001 Rear Wheel 3453922 1459395 237 00199 Effect Tests Source Nparm DF Sum of Squares F Ratio Prob gt F All Wheel 1 1 90423720 320500 lt0001 Rear Wheel 1 1 15802813 56012 00199 Analysis of Residuals H ig hway M PG ReSIdual 0 U1 I Residual Count Normal Quantile Plot Getting Started with JMP at ISU 1 Introduction IMP pronounced like jump7 is the new campuswide stande statistical package for introductory atistics courses at Iowa State University IMP d st starte 77 document will cover the options 0 Use IMP7s builtiin helpim the Index tab of the IMP contextisensitive help for most of the items IMP displaysijust select the an oniscreen item 0 The complete I MP documentation is also available in the same folder as The documentation is provided as searchable and printable PDF les a is produced by the SAS Institute This etting basics related to using IMP For further help you have several ost windows have a Help button or you can access the help using Starter window The tool can also be used to access the tool and click the IMP application 0 You can also email your questions to jmphepoiastateiedu 2 The JMP Starter window When you launch IMP it displays the IMP Starter window unless this default has been turned o If the IMP Starter window is n how1ng you can select it from IMP 7s Window menu The I P Starter consists of several tabs shown at the top that provide easy access to many of IMP7s features f Huuwarme r Hmmm r nmm f u mummmmaimmi a Yahleaan New We mug mm a new um um and mavlay a a and m a new wiryan m um mm um 5 MP me cnntaimm a datatahl2nrimvnrtit1rnm summer we NEW 5mm Write and mm m mcmamq mm Wequot 5mm um a me cnntaimm m a mm m examvle mmquot mm ummecmmmmqmmmampnmnm ExamineandaetayatemDr erence The Index tab provides help for all of JMP7s features which are listed alphabetically like a book index Selecting a topic and click the Topic Help button to access JMP7s help le for that topic or click the Run Script button to have JMP take you to the appropriate analysis dialog to access the selected feature The File and Tables tabs mimic the File and Tables menus respectively The remaining tabs group JMP7s features according to statistical topic You can access the IMP Starter window items in the menus however the IMP Starter organizes items in a useful way 3 The JMP Data Table FilegtNew creates a new data table in which you can enter data Each column corresponds to a variable and each row corresponds to an observation on each variable FileHOpen produces JMP7s Open dialog box where you can select les on your hard drive to open in IMP JMP will open several le types in addition to its own special JMP Data Table f rmat For example JMP will open SAS Datasets and text les containing data If you have data saved in a text le you can control the way the data is opened in J MP by having I MP display only Text documents in the Open dialog box and specifying that JMP open the le as Text Data with Preview In the following screen shot the data is in the text le called students txt This same data is available in JMP7s Sample Data folder as studentsjmp7 In order to have JMP provide a preview of how the data will be imported you must specify Text documents and Text Data with Preview in this dialog box EM m sum Iextmuuments up As m m wlth Pravlew In the preview window you can specify how the columns and rows are delimited in the text le For example this le uses commas to separate values so Comma is chosen in the upper rightihand corner of the preview window as the End of Field delimiter The column names and the rst two rows of data are displayed as they will be imported given the speci cations you7ve made in the preview window If the data doesn7t look correct here try choosing di erent delimiters until the rst two rows look correct Click the OK button when you are satis ed with how the data will be imported The above screen shot shows the IMP data table created from the imported text le The leftihand side of the data table window provides information about the columns and rows Each small red downwardipointing arrow E is a popup menu providing access to commands from the Rows Cols and Tables menus Each lumn has a Column lnfo dialog in which you can specify the name of the column its data type its modeling type and other characteristics speci c to that variable You can access the Column lnfo dialog via the Cols menu or by simply doubleclicking on a column title iFnrmat ism vwmm w You can also assign each column a prede ned role for the analyses you plan to perform You can specify that a column be an independent variable or factor by assigning it the role of X a dependent variable or response by assigning it the role of Y or you can specify that a column contains observation weights or frequencies You must specify data types continuous nominal ordinal for each column however you can leave roles unspeci ed if you are unsure of what role each column will play in the analyses you have planned You can specify roles on an analysisbyianalysis basis 4 J MP Menus JMP s menus are as follows File opening saving printing and the Quit command Edit cut copy paste Journal Layout Preferences Tables table manipulations sort stack split subset etc Rows row actions label hide exclude delete etc Cols column actions new delete column info roles etc DOE Design of Experiments features Analyze starting point for most statistical analyses distributions and models Graph every analysis platform in the Analyze menu provides graphics but this menu provides access to more graphs that JMP can create Control Charts are found here as they are primarily a graphical technique Tools Palette of cursor tools selecting an item here changes the basic function of the mouse cursor most of these items are also found on JMP s oating tool palette found just below the menu names Window window actions redraw hide etc Tool Elements Window lets you customize the oat ing tool palette Help access to JMP s help les 5 JMP S M O JMP s method of operation77 consists of the following steps H Identify the data type of each column numeric character row state to Specify the modeling type of each column continuous nominal ordinal 03 Specify the analysis role of each column and then choose a general analysis type usually from the Analyze menu you may choose a type of analysis and then choose the roles of the columns if you prefer F In the new analysis window that J MP creates explore the output provided request additional output and save results via the pop up menus in the window denoted by little red downward pointing triangles next to variable names and journal any results of interest with the Journal command 0 Print andor save your journal window Journal windows can be saved in various formats including jrn JMP Journal html for the web and rtf Rich Text Format for word processors 6 Analyzing Distributions To analyze the distribution of a variable choose Distribution from the Analyze menu Specify the variables you wish to analyze Here sex and height are chosen Click OK Em sebum Enlumn min Rum JMP creates a new analysis window with default output displayed the defaults can be changed in the Preferences The output displayed depends on the modeling type of the variables Contini uous variables result in histograms quantiles and moments while other types result in bar charts and frequencies unenlstnmlslrlhutmn Elnlslrmullnns Elszx m r M mm 233 VEIEIEIEIEI 2 mm mm v1megm 1 v1 Quanllles v Hnmznls mums mm 72mm Mean mam 995 m m 5mm 389525 5m r m Hz 9 nun unwaswm mum Quartile 54mm lnwergSKMean suamu median munu H 233mm 25m warm sauna nus 55 swim nus minimum 5 nun 7 Notice that each variable has a popup menu El providing access to further graphical and numerical ies as well as statistical procedures like con dence intervals tests etc The whole window also has a popiup menu upper eftihand corner by the word Distribut39 39 ne t an e th overall display and save the entire analysis as a JMP Script Furthermore if you Controliclick Macintosh or righticlick Windows anywhere in any JMP analysis window y ill see a contextisensitive popup menu that will let you further specify how the output is displayed and what exactly is displayed The di erent JMP tools also perform di erent functions for each analysis window For example the Hand tool if will dynamically change the class size of the histogram if you select it and then hold the mouse button down on top of the histogram and move the cursor vertically All J window elements are dynamicallyilinked For example clicking on the bar for F under sex will shade the female observations in the height histogram and select all the female observations in the data table If you doubleclick the F bar JMP will create a new data table with only the female observations unenliJmnmlslrmutln Ellllslrmullnns Elszx m r EnuntAxis v1 nuanmas v1 Hnmznls mu us maximum 72 nun Mean 5 33mm 995 7 mm 359525 in r s ssuuu 5 E M EnuntAxis 97 5 as u 2 9n us ss nun upper 95 Mean 5 53354 75 us warm 54mm mar 95 Mean su 527m so us u 233 nun median m nun 25m warm sauna m us as 2 5 53 nun u 5 5 nun u us mum at nun You can print the analysis window directly use the EditgtJournal command to journal the output or use the EditgtLayout command to create a live journal A JMP Journal window provides a word processorifriendly version of the output which you can save as a JMP Journal le an RTF le can be read by most word processors or an HTML le The Layout command creates a J MP Journal window in which the analysis elements can be rearranged before saving 7 Analyzing Relationships Between Two Variables The AnalyzegtFit Y by X menu command provides access to four di erent types of analyses de pending on the modeling types and analysis roles you specify for your columns The following table shows what type of analysis JMP does depending on your variable types and roles Y1 Xgt Continuous NominalOrdinal Continuous Re essionquot On 7 ay ANOVA NominalOrdinal Logistic Regression Contingency Table Analysis J MP also provides Spline Fitting options in this case The following screen shot shows the Fit Y by X dialog box The model between height contin7 uous dependent variable and sex nominal independent variable is being speci ed Clicking generates an analysis window with appropriate graphics and statistics From this window you can request further graphics and statistics as well as save information like residuals to the data table and journal the output for future use 8 Analyzing Relationships Among Several Variables The AnalyzeHFit Model provides a dialog for specifying complex linear models For example in the following screen shot a model involving age sex and weight and some interactions terms is being speci ed with height as the response variable One can also specify nested factors pmmmy mummies quare7 E AurmmZEl I n immem 7 The Fit Model dialog has ve popiup menus which allow for further details to be speci ed about how the model is to be t and what output you prefer to see Model Speci cation Personality Emphasis Stznwlse Ma nnva inunueauVaname Nnmmal inmsuc am llnIJI u P mm m Paramelrlz SunIva a I Attribut es Randnm mm Pnlynnmlallu name Exrluden mm As with all of JMP7s analysis platforms once you specify the model and click OK JMP creates a new window containing the default output for your analysis From this window popiup menus rovide access to further graphics and statistics as Well as the ability to save residuals tted values etc for future analysis You can also journal the output 9 JMP S Graph Menu While every analysis platfrom provides access to appropriate graphs for the analysis requested lMP provides a Graph menu to allow more exibility in What you are able to present graphically hm erlny Plnl nllll39nl hart arlahlllly hart arelu hart aw Pln rn ler nnluurPrn ler The Cha1t command provides bar line and pie charts The Overlay Plot command provides time series plots and scatterplots Note Scatterplot matrices are provided under the Multivariate command of the Analyze menu The Spinning Plot command provides live 3D plots that you can make spinihence the name Control Charts are found under the Graph menu as t e primarily a graphical technique Pareto Charts are also available here along with several other graphic options 10 JMP Resources 0 Use lMP7s builtiin helpithis may be the quickest Way to answer a question about lMP7s abi 39ties 0 Access the full documentation for JMP which is provided as PDF les installed with the application 0 The lSU Statistics Department maintains a support page for lMP at http www public iastate edu statcompguidejmp html 0 Some books about using lMP 7 JMP Start Stattsttcs by Sall Lehman and Creighton Published by Duxbury ISBN 0534359671 Data Analysts mth JMP N 4 by Johnson and Berk Published by Duxbury ISBN 053437395X P IN is the student Version of lMP o The SAS Institute has a Web page for lMP at http wwwl JMPdiscovery com
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'