Tosmana
Contents
1. 69 3 73 1 Thresholds 70 71 35 72 25 4 Missing 0 Median 71 65 M show min max values Decimals 2 v show threshold oem v show median In the central panel the distribution of data is shown Each orange box represents one case The green lines represent the thresholds If the option is set lower left corner the values the values of the thresholds are noted above the threshold line The number of decimals shown can also be set 4 The thick red line is the median the line can be toggled on off with the 10 option in the lower left corner 3 To select a different variable please use the dropdown menu at the upper left corner of the window 1 To move the threshold simply place the mouse on the wanted threshold the mouse pointer will change and press the left mouse button The move the line by moving the mouse still pressing the left button and release the mouse button with the value wished The number of thresholds used with the variable can be selected with the dropdown list in the upper right corner 2 By clicking the cluster button a simple clustering average linkage will be performed on the variable data Recode Stepped Conditions To recode a stepped condition into a new condition select Tools gt Recode Condition from the menu bar Select the condition to be recoded and check the proposed name of the new condition By selecting the old values found in the left box2. Updates A mailing list has been set up to notice about TOSMANA updates please see http Awww tosmana net joinlist html 17 8 Citing Technical and contact information If you use Tosmana please cite like this Cronqvist Lasse 2005 Tosmana Tool for Small N Analysis Version 1 25 Marburg http www tosmana net Version of Tosmana covered by this manual 1 25 Date of Software Release 6 January 2006 Tosmana is developed by Lasse Cronqvist Email cronqvis staff uni marburg de Postal address Institute of Political Science University of Marburg 35032 Marburg eC CC Ce Software Requirements NET Framework v 1 0 or higher Download of Tosmana http www tosmana net 18
3. Select CSV File csv as File Type in the save dialog window 4 Type the filename to be used save the file 5 Close the worksheet In Tosmana In Tosmana select File gt Import gt Excel and select the file just saved The import feature of Tosmana recognises the variable type case descriptors interval scale ordinal scale You can review these settings and then the data appear in the Tosmana Main Window You should then save the data as a Tosmana XML file Importing data from SPSS Also SPSS dat files can be imported with File gt Import gt SPSS Data dat As with Excel imports you first have to assure yourself that a variable is included which can be used as Case ID for your data set Please also notice that all variables are considered to be of a nominal type You may change this in the variable settings section see below To import data from SPSS to Tosmana please follow these instructions In SPSS 1 Open the data set in SPSS 2 Select File gt Save As Select Tab Limited dat as File Type in the save dialog window 3 Type the filename to be used 4 Save the file In Tosmana In Tosmana select File gt Import gt SPSS Data dat and select the file just saved The data should now appear in the Tosmana Main Window You should then save the data as a Tosmana XML file Importing from fs qca By selecting File gt Import gt Excel or fs qca file a DAT File saved by fs qca can b
4. TeTteeistrationcand sdppoEi coc eod Sae utet Do tr cH Asin eet cout et T 17 8 Citing Technical and contact information esssssssssssseseeeeeeee nennen 18 1 Introduction Within this document the reader will find a short guide to get started with the Tool for Small N Analysis Tosmana Although the basic features of QCA the main method implemented in Tosmana are described this is not a general introduction to this method QCA was first introduced to the social sciences by Charles Ragin with his book The Comparative Method where he sought to move beyond the qualitative quantitative gap by introducing a new method to combine the strengths and to reduce the problems of either approach This resulted in Qualitative Comparative Analysis QCA The main goal of QCA is to reduce the complexity of data sets by using Boolean algebra Since 1987 several software programs have been created to implement this method In every new program several things had changed New algorithms have been introduced and more powerful computers allowed extending the limitations regarding the number of conditions to be treated simultaneously With fs qca Fuzzy Set Qualitative Comparative Analysis a graphic user interface windows style was used for the first time and the 12 variable limit of QCA3 was overcome Tosmana also implements classical Boolean algebra but it seeks to tackle one of the main limitations of QCA its restriction to Boolean sets Ev
5. calculated by selecting the two expressions from the expression list using the shift key to select multiple entries and then pressing the computer intersection button The complement of a logical expression can be calculated by selecting the expression from the list and pressing the compute complement button From version 1 25 logical expressions can be directly added from the TOSMANA report please see page 15 on this 7 Registration and support Tosmana is free software and it comes as it is There is no guarantee for correctness of any part of the program Also Tosmana comes to you on a something for something basis The software is provided to you by the author but you also should report errors or new ideas for the software so that the software can be improved To do so please visit the Tosmana FAQ on http www tosmana net faq html Registration Tosmana is free but we would like to know who uses Tosmana Therefore we request you to register using the form online on http register tosmana net As a registered user you will be informed of changes in Tosmana and you will be able to ask for support on Tosmana if you have any questions on how to use it You will receive a registration code when registered which can be entered into Tosmana Support If you have any trouble using 7osmana you may ask for support on the Tosmana web site Please use the web forms provided there as these are more easily processed than emails
6. in which data files are created loaded and saved and from where the calculation window is called In this chapter the features of the main window are described When you start Tosmana you are asked if you want start with a new file or load an existing data set If you which to load a file previously edited with Tosmana select Load a data set In any other case please select Empty data set If you would like to start with an import from excel or SPSS please select Empty data set and then import the file from the main window Loading and Importing Data Loading a file into Tosmana A file created and saved by Tosmana can be loaded by selecting File gt Open from the menu bar in the main window Please note that only files created with Tosmana and saved in the XML extended mark up language file format can be loaded with the Open item Please refer below on how to import files from other programs Excel SPSS Importing from Excel By selecting File gt Import gt Excel a CSV File saved by Excel can be imported to Tosmana This feature has been tested with Excel97 and Excel2000 but it should work with other releases as well First you have to assure yourself that a variable is included which can be used as Case ID for your data set To import data from Excel to Tosmana please follow these instructions In Excel 1 Open the data set in Excel Switch to the sheet which is going to be exported to Tosmana 2 3 Select File gt Save As
7. 1 Where applicable Windows2000 NT XP You have to be logged in as an administrator on your system to be able to install the framework 2 Download the NET framework onto your system A link is found on the Tosmana site 3 Execute the file dotnetredist exe just added to your system Please read and accept the end user licence agreement EULA of the NET Framework by Microsoft and then choose a directory to extract the installer Please note where you placed the installer 4 After the installer has been extracted you can install the framework by starting the installer dotnetfx exe from the windows explorer 5 If you are asked to restart your computer do so 6 If the installation was successful you can now install the Tosmana software Please note that you will only have to install the framework once When you update the Tosmana software at a later stage you will not have to reinstall the framework Installing the Tosmana software With the VET Framework installed you can download the current version of Tosmana as a zip file from the site mentioned above Please extract the two files included in the zipfile to a directory of your choice Please ensure that both files are extracted 3 Constructing your data set If the VET Framework is installed correctly the Tosmana preview can be started easily by running it out of the windows explorer like any other Windows program When Tosmana is started the main window shows up
8. Lasse Crongvist Email cronqvis staff uni marburg de Tosmana TOol for SMAII N Analysis version 1 25 User Manual Release 6 of January 2006 Content Ts Introduetioticsoo ideft auth stu eto rare tr rep egeta a me eee ee eer a eM Ld abo E 3 2 Installitig Cosmana sty acca erus ese tot een iu ERROR Testa dilosu didus READS en as Mea bend 4 Installing the INED Frameworks 8 oi eoe denon De yp eub o oaa eate pe un De pe tuna e 4 Installing the Tosmana SOTCWAEG vss icri roce dL tke Eee ep ete ue UR TE Gar ee Ee Ue e e RR dd 5 3c Constr ting VOUT data Seb us cereo rabie curd bred re eua ore eed o tius 6 Loading and Importing Date cetus aed e a ea lesion cadet abu eus Bee ids 6 Crest and saving Tilesiuce ete o pa reed Reuben Piece aed 8 Editing the dati Sel cun duae GN oat edo a Rua eR e eb aae sb to vut ul 8 Adding and editing variables 2e M eere qr co i eee eive eaae eL cane I RENE E MERE E E E egi 8 Transforming Scale Datarea tese T aet GI OPo UD E ane acest dva AU tina ACE 9 Recode Stepped Conditions 4i oett arr d etre etis vet es die eoe o ieu RAS to lu esent 11 4 Data UE NT 12 Manual Selection of Prime Implicants uui eece tens tope tego Cert teneor Env edlteb eset pr eb eed esae 14 The Tosmana Reporte eet aca dae bale ep Ls e ae Cedo Y dig 15 S T he ISO ZO oer ee ERR leue ea da Men URDU NURSE we ra iN Dun 16 Oy Phe Boolean Calculator coe teet sebo dedi ea ttis eun uat fos 17
9. e imported to Tosmana First you have to assure yourself that a variable is included which can be used as Case ID for your data set Please also notice that all variables are considered to be of a nominal type You may change this in the variable settings section see below To import data from fs qca to Tosmana please follow these instructions In Excel 1 Open the data set in fs qca 2 Select File gt Save As Select comma as File Type in the save dialog window 3 Type the filename to be used 4 Save the file In Tosmana In Tosmana select File gt Import gt Excel or fs qca file and select the file just saved The data should now appear in the Tosmana Main Window You should then save the data as a Tosmana XML file Creating and saving files By selecting File gt New a new Tosmana data set will be created You can add variables to the data set by selecting Dataset gt Enter new Variable from the menu bar As soon as all variables are added you can add the cases of the data set simply by entering them in the data matrix of the main window To change the Variable name you have to enter the variable edit window described below Selecting File gt Save in the main menu bar saves the data set If you wish to use your data set in other applications Excel or fs qca you may also save the file as a csv or dat file choose Export Editing the data set You might change the content of the data matrix at any time by c
10. ery element of a data set has to be T true or F false but often we want to use more condition values like ow medium high Therefore Tosmana introduces Multi Value Minimization as an additional feature of Boolean Minimization but it also can be used for Boolean Minimization As this is a software manual the method as such is not discussed here but the user should read available introductions to Multi Value Minimization Please also visit the Tosmana resource site where a list of further readings is found As Tosmana is work in progress many details of the software are subject to changes Ragin C C 1987 The Comparative Method Moving Beyond Qualitative and Quantitative Strategies University of California Press Berkeley Cronqvist Lasse 2003 Using Multi Value Logic Syntheses in Social Science Paper presented at the 2nd General Conference of the European Consortium for Political Research ECPR Marburg Found on http www tosmana net resources 3 2 Installing Tosmana Tosmana is using the NET Framework a programming platform provided by Microsoft The advantage of using this framework is that it takes care of many details of the software creation so that the programmer can concentrate on the central topics This means that creating Tosmana is much faster with NET than without a programming framework But this also means that the user has to install a version of the NET Framework if there is no such f
11. hanging the data in the main window Please note that you have to open a new calculation window to use the new data in your calculation If you want to add a new row case just enter the data in the last row of the data set If you wish to delete a case click on the row header of the case you wish to delete and press the delete button on your keyboard Adding and editing variables To add a new variable please select Dataset gt Enter new variable Edit the name of the new variable and press ok The new variable will then appear in the data set If you wish to change the attributes of an existing variable please change to Variable View by selecting Edit gt Variables Settings e To change the name of the variable click on the button labelled change name and enter the new variable name e To remove a variable entirely select the variable name in the Variable Settings window and press the button labelled remove You may add a description to a variable in the description field Ordinal Nominal data If you just work on Boolean data sets you do not have to think about if your variable is of a nominal or ordinal type but when you are using Multi Value data sets you should be aware of the differences in calculation with those different types of variables see chapter on set notation below Replacing multiple data entries If you wish to replace multiple data cell entries with the same value you may do so using the replace
12. ical data as outcome or condition an error window will show up Visualizing your data Up to five conditions can be displayed in a graph by pressing the visualize button 2 See the chapter on the Visualizer below for details Selecting the calculation style Tosmana provides you with three different ways to calculate solutions which can be selected with the Selection Mode drop down box 3 12 1 All solutions are calculated and displayed in the result report Show All 2 When more than one solution is possible the user can select the prime implicants used in the manual selection window see below but Tosmana ensures that the solution is minimal Select if possible 3 The user may combine the solution freely in the manual selection window but this means that the solution does not have to be minimal Free selection Calculation configuration The configuration of the calculation can be set in the right half of the MV OCA window 4 The configuration labels are to be understood as follows e Explain Those outcomes which are explained by the solution in QCA2 Matrix to be minimized e Include for reduction Outcomes allowed to be included in the reduction of the implicants in QCA2 Matrix for computing Prime Implicants please note that outcomes to be explained had to be included in this matrix in QCA2 in QCA3 don t cares e Exclude Outcomes which are excluded from minimization Select Notation Regu
13. larly Tosmana shows the expressions in the solution with strict set notation If the option Use upper lowercase notation is checked the notation used with QCA2 and QCA3 can be used instead Show Implicants To only get a list of the prime implicants press the show implicants 6 button In the prime implicant window coming up you can then see the prime implicants for your selection of conditions and settings You may filter the list by cases With a click on a prime implicant in the list the cases implied by this implicant are shown Starting the calculation Start the calculation with the Go button 5 The result report will appear in a new window of the Internet Explorer This window might be saved or printed 13 Manual Selection of Prime Implicants In the manual selection window prime implicants are selected If there is a possible choice between prime implicants to find the minimal expression these are displayed in the lower part of the window whilst already selected implicants are displayed in the upper half These are either selected by the user or by the software indicating that these prime implicants have to be included in a minimal solution Cases not represented by already selected implicants have green colored columns to indicate this To select an implicant highlight it by selecting the row of the implicant with a mouse click and press the use selected implicant button You may press the reset button to undo your selectio
14. ment window select Edit gt Replace Define the search value and the new value to replace it and select all variable where this replacement is done Press ok and the replacement will take place You can use this feature to change the missing data value in the data set to the needed in TOSMANA all at once Variable transformation The menu item Transformation in the Variable Settings Window defines how the raw data is converted to the data used by MVQCA There are four different options which can be set for each variable use original values 1 1 Using this option indicates that the raw data is used as it is with MVQCA If you are importing a data set from SPSS or you are creating a new data all variables will be set to this option Please note that all data must be of numerical type and that missing data always has to be noted with an Variables containing the name of the cases should be set as use as case ID If this is not the case Tosmana will produce errors no Transformation possible If this option is set the variable can not be transformed and it can not be used with the software The sense of this is to exclude variables which can not be translated by the standard rules yet but which may be used in future like calendar dates etc use as case ID If this variable is set the variable can only be used as case ID use Thresholds If this variable is set you can use the threshold setter described below to set the thresh
15. n s When all cases are covered by the selected implicants the result report is called by pressing the ok button The cancel button closes the manual selection window without creating a report Select Prime Implicants Selected Implicants ok Term 34 AUS GER ITA HUN ROM POR SPA GRE k LITERACY40 Prime Implicants not used HUN ROM POR SPA GRE POPULATHOYLANDLORD41 POPULATI 1 5EAWARD40 POPULATH1 LANDWARD4 1 POPULATI 1 EARLYSTA 0 POPULATI 1 PREWARDE 0 POPULATI 14NATPRODC 0 _ POPULATI 1 URBANIZA O SEAWARD D LANDLORD11 LANDWARD 1 LANDLORD41 NONREF 1 LANDLORD 1 REFORM D LANDLORD11 EARLYSTA D LANDLORD41 PREWARDE 0 LANDLORD 1 URBANIZA O LANDLORD 1 14 The Tosmana Report When the calculation is finished the result report opens in the Tosmana result window or if selected in the Internet Explorer Results File Options Tosmana Report Algorithm Graph based Agent File C Dokumente und Einstellungen lasse Eigene Dateien uni KURSE MARBURG psvergleich msdoue xml Settings Minimizing Value including Truth Table vi URBANIZA 2 INDLAB v3 NATPRODC LITERACY OUTCOME id COUNTRY 2 id SWE FIN IRE HUN EST POL BEL NET FRA UK CZE GER AUS ITA ROM POR SPA GRE Results alD INDLAB O NATPRODC 1 SWE Created with Tosmana Version 1 202 You will find the settings
16. of the calculation and if selected the data matrix first in the report At the end the solution and if the option is set the simplifying assumptions are reported Unlike in earlier QCA Software Tosmana can either use upper lowercase notation to display the results or use strict set notation As Tosmana also allows more than two possible values for each condition the upper lowercase solution can only be used to display strictly dichotomous data sets Any parts of the report can be copied to the clipboard and pasted to other application by clicking the right mouse button and selecting copy in the popup menu Also solutions can be directly added to the Boolean Calculator by highlighting the solution and selecting Add to Boolean Calculator from the context menu pressing the right mouse button gt Please see Charles C Ragin 1987 The Comparative Method Moving Beyond Qualitative and Quantitative Strategies for the original detailed description of using Boolean sets in social science and Gis le DeMeur Benoit Rihoux 2002 L analyse quali quantitive compar e AQQC QCA approche techniques et applications en sciences humaines for a introduction how to apply the method The use of Set Notation is explained in Cronqvist Lasse 2003 Using Multi Value Logic Synthesis in Social Science Draft found on http www tosmana net resources Currently this only works with dichotomous data sets The use of set notation in the re
17. olds dynamically while working with TOSMANA Please note that all data must be of numerical type and that missing data always has to be noted with a Transforming Scale Data To transform scale data to stepped conditions suitable for MVQCA analysis thresholds must be set to recode the data TOSMANA provides two ways to do so Thresholds can be set directly by hand or thresholds can be set interactively by using the Thresholdssetter To set the thresholds directly choose Edit gt Variable Settings from the menu bar and select the condition to set the thresholds double click the row header If the option use thresholds is selected as transformation mode the button edit can be pressed The thresholds are entered by using dot to separate decimals and to separate multiple thresholds Using the Thresholdsetter Window If one or more Variables are containing data which is scaled on an interval scale the data must be transformed to an ordinal scale before it can be used with MVQCA To do this the variable must be set to be transformed with threshold s in the Variable Settings window set the Transformation item to use Thresholds Then the Threshold Settings Window can be opened with Tools gt Thresholdsetter in the main window Please note that only missing data noted by will not be shown in the threshold setter ThresholdSetter ojx atsliable DALE x gt Thresholds 3 cluster
18. port is required unselect the Use Uppercase Lowercase notation option in the MVQCA window 15 5 The Visualizer Tosmana allows the user to display a Boolean data set with up to five independent conditions to be displayed graphically To do this select the conditions in the MV QCA window and press the visualize button You will then get the Visualizer window Visualizer el Fa Highlight Cube Highlight Solution NONE v Show labels v Short variable names save dispose Black White Please note that at the recent stage it is only possible to save the created graph as a bitmap file which is rather large but this can be edited and compressed in almost every graphic suite In the Visualizer you can toggle between Colour and Black White display and select which information is shown You might also highlight solutions of your choice 16 6 The Boolean Calculator The Boolean Calculator allows simple calculation on Boolean Expressions like intersections and complements To perform such a calculation the Boolean Expressions have to be entered into the calculator This is done in prefix notation just like using a standard pocket calculator First the single expressions are entered which then are connected by and When the logical expression is entered it can be added to the lower calculation box by pressing the add button An intersection of two logical expressions is
19. ramework already installed on the system In this chapter the reader will therefore first find an instruction on how to install the NET Framework and then an instruction on how to install Tosmana If you have trouble installing any of the parts mentioned you should contact the Tosmana FAQ on the Tosmana resource site where you also are able to contact the programmer see chapter 6 The Tosmana resource site is found at the following URL Tosmana resource site http www tosmana net Installing the NET Framework Tosmana requires the NET Framework to be installed on your computer to be able to run The Framework works with all recent versions of Windows Windows98 and later and is distributed by Microsoft for free and can be downloaded from the Microsoft site On the Tosmana resource site there is a direct link to the NET Framework download page on which you find detailed information on software requirements and installing instructions for the framework Below you will find a short step by step instruction for installing the NET Framework that should be sufficient for most users The framework comes as one large file 20 MB which has to be downloaded before starting installation The framework also requires some other software installed on the system please see the exact description on this and how to install it on the Microsoft site follow the link on the Tosmana download site Installing the NET Framework SDK step by step
20. these might either be used unchanged in the new condition button or assigned new values enter the new value in the text field under the Convert to button and press this button If you wish to undo a coding select the recode expression in the right box and press the undo button 11 4 Data Analysis The main goal of Tosmana is to perform minimization on Boolean and on Multi Value Data sets This is done from the analysis window named MV QCA Call the analysis window from the menu bar Tools gt Start MV QCA MV QCA Case Descriptor COUNTRY dependent variable DUTCOME Explain Max extra length reduction Exclude independent variables 1 Outcome 0 4 e C C Dutcome 1 C e Lontradictions C G I Missing Outcome C C I Remainders s LL e Ignore one case implicants v Show Data Matrix in Output Iv Include variable information in Qutput Close Reports on Exit v I Compute eue Assumptions 6 show implicants PAROCHIA Figure 4 1 The MV QCA Window Selecting variables for analysis Before the calculation can be done the variables for the analysis have to be chosen red box 1 in figure 4 1 Use the arrow buttons to select the outcome and the conditions Please note that the use of a case descriptor is necessary in 7osmana Variable values have to be of the type 0 1 for Boolean analysis and 0 n for multi value analysis If you choose a variable with non numer
Download Pdf Manuals
Related Search