Qualitative Research For Beginners
Contents
1. aoe Marbionald Farm rif gt ss 1 Cusi ir Wortbareh Cede srar a name acl Ae arm rit verse 1 Cd Masdasnakl had a larn EIEIO and on his farm he hed a pig EIEIO a wilh an Gin GMR here and an oink oink there here an oink ertan oink erenywhere an oink oink Old Macdonald had Sond gt pig jn 1 Select the relevant text Sa May e g SL Pate l Fillin the new code here 3 Click the New button Figure 9 Creating a new code Then click the New button This s the one in the button panel A dialogue will drop down asking for the code s definition 19 MacDonaldsFarm rtf Describe code sound gt pig Color The sound children attribute to pigs Mi use inherited color Go with an oink oink here and an oink oink there Figure 10 Adding a Definition Clicking the Ok button will have 3 effects 1 This code and definition will be added to the code list in the workbench 2 The code list on the left of your document window now has a new code in it sound gt pig and 3 oink oink has now been coded 20 Kacai lda Farr mi E 5 A fi Easicrize Kar Barkkareh Fee Search WMacDonaldasFarrn rtl yerse I Old Hacdonakd had a larm ElEW B and on his farm he had a pig EIEIO Ei wiih an soundepid Psoundspig here and an aink GNE there z here an oink here an ginik Bverwfiere an oink oink e er Be Old Maedonald had a farm EIEIO2. It also shows some of the issues that can result from such searches For this example I m going to use this small passage R Well my high school was known as a trouble school There were a lot of fights and kids uhm wandering around and most of us worked in factories on the city s east side Most of us partied rather than worked We ll code this as follows R school gt trouble Well my high school was known as a trouble school aspirations We weren t going anywhere aspirations violence There were a lot of fights violence and truancy kids uhm wandering around truancy and aspirations most of us worked in factories on the city s east side aspirations gratification gt delayed Most of us partied rather than worked gratification gt delayed school gt trouble Now let s imagine that we are searching for the intersection of discourses we ve labeled school gt trouble and aspirations Maybe we want to see how high school image and occupational aspirations correlate One way to find these intersections is through searching for school gt trouble aspirations You can also find this from an unlimited search however Reasons for doing so include the ability to do set mathematics 1 e the ability to look at how groups of coded passages intersect with each other To do this we re going to work from an unlimited search that means we hit the search button with no criteria filled in
3. Let s enter a second code Click the Save Clear button and add a code sound gt dog with the definition Marks text showing what dogs sound like to children 17 When you are done typing the definition either click a different tab or press the Save Clear button The program will save the definition in the project window Adding a new code from the document window Let s code Old MacDonald s farm or at least one verse so that we get the idea This reiterates what we talked about initially with the New button Make sure you have an open document window If you don t already Click the New button on the workbench This is the new button you should be looking for not the one in the code buttons _ Ne f Open j f Remove f Add Sets w Files Init File Add fa Add all gt gt u lt Remove lt lt Remove al Figure 8 The new button to make a new window This should provide a blank window Type the pig verse into the window 18 Verse 1 Old Macdonald had a farm EIEIO and on his farm he had a pig EIEIO with an oink oink here and an oink oink there here an oink there an oink everywhere an oink oink Old Macdonald had a farm EIEIO This will go on the right hand side Let s now assign the code sound gt pig to the oink oink Select oink oink and then on the left hand side type in sound gt pig
4. conversion tools exist As for multimedia files they need to be converted to some format that Quicktime supports AIFF WAV MP3 etc The first phase of my projects involve frequent scene shifts between the field and the computer At this point in the research cycle I m using TA to transcribe interviews TA has a little transcription machine built into each document window or to make logs of video I ve also entered any codes 1 e words or short phrases that name themes up front I m also adding codes as I see themes emerge while transcribing There is another type of decision I m making as I enter codes at this point This concerns the types of independent variables that I will use later on in analysis These include who is being interviewed observed the times of particular observations or points in the interview aka time codes SES geography gender race or other sociological factors I m l I have been using the command line version of MPlayer which is available through the Fink project I type mplayer ao pcm myfile wma and a wav file is generated from the wma interested in Some of these should be done as data is entered the time codes for instance It s just easier that way and TA has some tools that automate the process Others I can enter later though if I know that a particular type of data is going to be of interest I can enter it as I go along TA has ways of saying in this file the gender
5. mark the examples of name calling and recode them as negativeBehavior gt nameCalling The gt is TA s way of indicating that nameCalling is a Subtype or subcode of negativeBehavior You can add subcodes to subcodes by just naming your codes with another level of the gt sign You would then continue marking the next type of negative behavior throwingSpitWads and doing the same thing Eventually negativeBehavior turns into a whole family of types of negativeBehavior all under a top level with that name The real challenge here is to try to categorize each instance that you come across of your top level code of interest Usually some are too vague to be classified and are thrown into miscellaneous subcode but you want to limit the number of these if possible For some projects this may be enough You have now created a typology for the themes you re studying By repeated reanalysis you can add layers of nuance to each code You can also create very sophisticated counts of these codes and subcodes for reports The counts are designed to be exported to Excel and to be analyzed using pivot tables and other advanced features However for many projects there is a need or desire to think big again or to think outside of your initial coding scheme and that is where phase 3 begins III Phase 3 From deep to broad After breaking down codes into subcodes and subcodes of subcodes you are ready for the re
6. and graphing of results but TA can take you at least 4 5 of the way there For many simple projects it may be all you need Starting TA and understanding the parts When you first run TA you will see the New project creator Create new project Project name Untitled Project location Users matthew Browse Add to work menu Work p Cancel Create Figure 1 New Project Creator Using this you pick the location and give the name to the new project Do not use any extension when entering the name 1 For this tutorial call it Children s songs 2 Before you click create determine where tams should save this information by using the browse button to pick the folder where you want your project saved Finally because this is a file we will work with throughout the tutorial let s add the project to the work menu 3 Click the Add to work menu check and 4 then click the create button You will notice that you have one very complicated window It is called the workbench or project window Every research project you have has to have one of these The file associated with this window it has the extension xtprj for XML Tams Project contains project wide information like the location of your interviews and other files as well as the codes which will be explained later and their definitions Note that the Project window will save itself as changes are made A The Workbench window T
7. at what sounds are associated with what verses I ll need to exit the dialogue and select all of the sound codes from the table Hint from the results menu pick select all and then pick Results gt Select and fill in sound Again pick Reports gt Data comparison table Again leave the top as codes pick versus for the vertical dimension area pointed to by box 3 and pick data again from the table pointed to by box 4 Click Show report verse compared against Codes verse sound gt cat sound gt pig 1 oink oink oink oink oink oink 2 Imeow meow meow meow Figure 12 Results from data comparison That should allow you to get going A couple of hints if you wish the codes to go down the side and the verses to go across the top check the Switch axes switch Indicated by box 1 in figure 10 Also you can select multiple fields to display in the table by shift clicking and option clicking items in the Box 4 table You can also rearrange the list in 39 the box 4 table by dragging items around You ll also want to explore what the count and grouping options pointed to by box 4 do to your table Have fun think sand box 8 A more complicated example Less for beginners Before concluding this tutorial I want to work through a more advanced example of selecting This is an alternative way of finding the intersection of coded passages by using Result gt Select
8. data Here s my guided tour of the document window In document windows there is a toolbar running across the top which is where you 12 Mail can put text and codes you use a lot The document window toolbar also has a button to move the workbench to the front and to save the current document Underneath this are two tabs like those for a rolodex One says edit the other says search We ll just work in Edit mode for now which is the one that the document window starts in The other tab is for searching for data in this particular file The Edit pane is divided for practical purposes into two halves the left side which has buttons and fields for managing codes well there s one button here which toggles the ruler but everything else is about codes on the left side of the window On the right side is a big pane which is where you can enter your text Note this was written before 2 41 when an additional set of buttons was added to the left side of the window While everything in this documentation holds you do have to overlook the media player a sort of built in transcription machine which has buttons labeled lt lt gt X and v To read more about the media player see the Audio Visual How To in the How Tos folder 13 6068 Untitled 2 Customize Save Workbench Edit Search Old Macdonald had a farm EIEIO Media audio and visual and on his farm he had a pig EIEIO control panel player w
9. is male or the city is NY etc so you don t have to mark every instance with that information You can also automate this information by telling TA that if Bob is the speaker he is from NYC is male Latino and bald Any time something is identified as from Bob that other information will be entered automatically The types of information I am talking about here I call context codes since they provide information about the context for the data they describe who the speaker is more than what s he said This identifying of context and data codes continues until I have collected the body of interviews and observations in my data set or until the interviews and observations are not generating new themes Then it s time to go deep II Phase 2 Going deep Phase 2 begins by reading and rereading your forgive the change of address data and checking and applying the themes throughout your database You are adding themes and every time you do you need to return to every document to see where those themes apply While not as physically taxing as transcription this is in some ways the most tedious work in a qualitative study The work ends when you ve stopped adding themes I ve found two tools in TA are absolutely essential in this part of the process The first is the definition button but as more codes 1 e themes are added this has to be supplemented by searches for particular themes codes Normall
10. of the document followed by the number of characters into the document that the passage started and finally the initials of the person who has checked out this document which only applies to multi researcher projects The columns that tams creates always start with the _ underscore character 32 Sorting is very easy at this point We can sort the data just by clicking on a header and pressing the sort button This is not the best way to sort information though a subject we will address in a future section C Searching for specifics There is a lot that can be done in unlimited searches and more often than not that s what I use to explore my data Often however the researcher wants to see only a single code or a few codes While that is possible to do in an unlimited search see the section on selecting data it s easier just to search specifically for the codes of interest The easiest way to search for a specific code is to simply double click the code from the list of codes on the workbench So if we want to see all of the sections of text coded as sound gt pig we could double click sound gt pig from the list and hit the search button Then only the sound pig records would appear in the results window To clear the search field click the button called Clear on the work bench 1 Double click from this 2 The code will automatically 3 Then press code list which you be entered as the search string the search want
11. to search for button A idia apri arch Dative coches City iali kiim Ed Ar Eri pty Exam Hen wrrpk Setting ural dund ial sound pig IT al T irar You can examine the definitions This area is used to of your codes by selecting the code modify the list of codes to those and clicking this button that you are most interested in Figure 4 Searching for a Single Code 33 You could have also just typed the name of the code into the box labeled Search What If you were looking for either sound gt cat or sound gt pig Simple Make sure that the Search box is empty by pressing the Clear button Double click sound gt cat off of the list Type in a comma and a space after sound gt cat in the Search field box Double click sound gt pig Press the search button ee Alternatively you could have typed in sound gt cat sound gt pig into the search box with the keyboard and then pressed the search button You could keep adding commas and different codes If you wanted to search for three codes you could type sound gt cat sound gt pig setting gt rural for instance The codes sound gt pig and sound gt cat are part of a family of codes called sound If we just look for sound we will get everything that is in the sound family including anything labeled sound sound gt pig sound gt cat sound gt cat gt persian etc If I only want to look for those things that are sound and not sound
12. to the search list those that TAMS should search If you want all of your documents brought over use the gt gt button If you want to empty the search list of all documents use the lt lt button Otherwise click on one file in the file list that s the left hand list at the bottom and click on gt to add it to the search list If you want to remove a file click on it in the search list and click the lt button You can also change the order of files to be searched by using the and v buttons on the left hand side of the screen 29 The File List The Search List ao 6 F OidMacK xtpry Files search etine cone ode sah ema i Mi m hakir me Seidel rut Fila a Gl init Tik Faith mode Buttons to add and remove files Buttons to shuffle the from the Search List order of files in the Search List Figure 1 Assembling A Search List On The Workbench in the Files Tab For our project included with this documentation we have one file in the file list We need to add that file oldMacD rtf to the search list so click on it on the left side of the project window and press the button labeled Add gt to add it to the list of files to be searched B Executing a search We have our search list now we need to tell tams what to search for This we do through the Search tab of the workbench window Our first search will be what is called an unlimited search meaning we re not pu
13. 40 gratification gt del school gt trouble Bi school gt trouble school gt trouble 2 aspirations school gt trouble 3 violence school gt trouble 4 truancy school gt trouble 5 aspirations school gt trouble l6 Figure 13 Picking _code Once we have done a search we are going to look for records that have both As a first cut let s work with just those records that have school gt trouble Select the _code column by clicking it s header and pick Results gt Select Fill in school gt trouble and hit ok select Select string school gt trouble r a i a Search case insensitive i be ool Floating point numbers a ool _ Regular expression 1l Figure 14 Selecting _code Now we need to find the records that are left that have aspirations in our example there is only one record to search but that just points to the artifice of the example Pick the _data column this time Then pick Results gt Select and enter aspirations As the search string voila what you see are records that have both codes 41 WARNING In our example we would get a different number of records if we first searched for aspirations in the _code column and then school gt trouble in the _data column We would eventually see the same text as we browsed through the records but you should know that the number of records would be different Finding how many passages coded with A ha
14. A provides two tools to help you catch these problems 1 Broken up codes sometimes the mouse slips and tags can end up in tags setting gt ru sound gt cat ral Here sound gt cat has accidentally been inserted inside of setting gt rural This will not make any sense to TA If you pick Check for pairs this will select problem tags basically tags that don t seem to have an end or beginning The one it shows you probably is not the problem tag but it will be near the problem tag It is a clue as to where the problem is For some reason TA can t find it s other end 2 Incomplete codes Sometimes in working with a document a tag at one end or the other will get deleted The solution is the same as for problem 1 Choose Check for pairs off of the Coding menu A tag will be selected if there are problems 1 e if there are not an even of beginning and ending tags This is a clue to the problem for some reason TA did not find a match for this 3 Nested codes Sometimes the same codes can end up inside each other This might be represented by the following situation a Some text a that I m a trying to code a 24 This is not the sort of nested code that works with TA It would be fine if the inner code was any code including a subcode of a if it were a gt b for instance The problem is that TA can t figure out where the passage ends and it will choose the shortest passage The phrase trying to co
15. Qualitative Analysis for Beginners using TAMS Analyzer a gt b qo SD Matthew Weinstein mweinste kent edu http tamsys sourceforge net 2005 by Matthew Weinstein all r ghts reserved Introduction This document provides an overview and tutorials for starting qualitative researchers using TAMS Analyzer TA The first chapter provides a birds eye view of TA and the way it fits into qualitative research It is a personal account of doing qualitative research and using TA to track data and answer research questions Other people may use the tools in TA very differently but this should provide at least one vision of how you might use this digital tool with your project To be clear the research I discuss in the abstract assumes a single researcher TA has multi user capabilities Using TA in a multi user environment is not simple and is discussed in a different document The point of part one is not to teach you how to do anything merely to outline the ways that TA is used to solve different problems at different stages in a research project The second chapter moves to specifics and talks about the way that coding works Coding is the process by which passages in documents are assigned meanings It also talks about setting up a project initially Coding is very flexible in TA A code 1 e a word indicating what a passage of a text means can mark O characters a few characters or a whole document The third
16. What is coding Coding is simply a way of transforming raw information into data In qualitative research 1 e research relying primarily on interviews observations document collection analysis proceeds by sifting through these raw or if transcribed semi raw pieces of information and deciding what each portion represents In even modest size projects this can produce what is known as a data burden 1 e too much information to be comfortably handled Coding is one way to handle this Using computers sticky notes or scribbles in margins and much more complex systems have been devised relevant passages are coded L e labeled as to what that passage represents Single chunks of text should be able to receive multiple codes and in most modern coding systems they can receive various refinements to those codes 1 e they can be subcoded this isn t just an example of X whatever that may be but a subtype of X called Y in TAMS the code would be X gt Y the gt symbol is used to indicate various levels of sub coding An example would help Consider a project involving studying the sounds that children think animals make We might begin to collect songs and rhymes that have animal sounds in them as well as interviews about animal sounds Of course Old MacDonald would be part of our data Take just this verse Old Macdonald had a farm EIEIO and on his farm he had a pig EIEIO with an oink oink here and an oink oink the
17. a farm EIEIO and on his farm he had a cat EIEIO with a sound gt cat meow meow sound gt cat here and a meow meow there here a meow there a meow everywhere a sound gt cat meow meow sound gt cat Old Macdonald had a farm EIEIOf setting gt rural Here the sound code sound gt cat is twice coded inside a stretch of setting gt rural Recalling the definition of a code When you have a lot of codes and sub codes pig is a subcode of sound in our example it is often hard to remember the definition you gave a code No problem Click one time on the code you re curious about from the code list and click the Def button There s one on the workbench and on your document window Your definition will POP Up 22 al WAGE Lr CALA ff sound gt goat Marks text showing what children think goat sounds should be ea i Figure 12 Code definition Working with codes As has been noted codes can be overlapped and nested Furthermore related codes can be nested and overlapped There are no problems for instance with fa This is my a gt b text la gt b la even though a gt b is a subtype of a That s fine From TAMS point of view they are as different as pickles and tomatoes That doesn t mean that codes can t cause trouble they certainly can See the section on Problems with codes There are several tools that make working with codes easy These are on the Coding menu Often while coding I f
18. aker speaker bob speaker identity this is me identity speaker mary speaker identity gt negation no it s not identity gt negation 2 Icontext speaker time time 10 time speaker bob speaker identity this is me identity speaker mary speaker identity gt negation no it s not identity gt negation time 225 time speaker bob speaker identity can prove it identity speaker mary speaker ok do so 2 What is analysis Analysis is a process of finding out what information is present in your data and what that information means Practically it is the process of taking codes and finding out what pattern they form and testing meanings for that pattern through searching for negative instances for instance To find these patterns in TAMS you ask the program to turn your interviews into tables that you can use to count instances of particular codes or collections of codes etc In essence what TAMS is all about is taking those codes and turning your interviews field notes etc into a database or spread sheet that you can 28 browse sort or further search n a variety of ways It s th s culling through the data which is the heart of analys s 3 Searching sorting selecting and comparing the grand plan At the beginning of your analysis after you have coded your documents there are three essential operations you need to master searching sorting and selecting In addition you need to
19. al analytic phase By now you have moved from raw data to cooked data data that signifies The third phase involves massive searching of your data and using a unique feature of TA searching searches winnowing down the results of searches until you find exactly the data you need and the pattern of codes in your data TA provides both search for codes and search for data string searches It also supports a very powerful standardized formatting language called regular expressions regex for specifying very complex lexical patterns Note that while TA provides tools to bookmark data search for significant patterns of code and text and to combine searches in a variety of ways e g to find the intersection of searches the real art of finding patterns within your data largely depends on your imagination Eventually you may come to find that families of codes also have relations to other families of codes or codes to codes not in the same family These non familial groupings of themes are called code sets in TA Code sets are simply arbitrary groupings of codes ones that are related and unrelated Code sets can be named with full sentences you don t have the no spaces restriction of code names In my work I have found that I use code sets to represent ideas that I want to test my well coded data against For example in a recent study of pre service teachers using science fiction to teach science content I started
20. and final chapter of this primer covers analysis the process of mining a document that has been coded for information Some of this information may be quantitative e g How many times did I use code X but most will be qualitative What are the examples of codes X Y and Z organized by gender of the interviewee These chapters are in no sense to be taken as a comprehensive description of TA or its capabilities When you have completed this introduction you can read through the user guide and release notes for descriptions of the full extent of TA s abilities Living the qualitative life in a digital age with TAMS Analyzer Matthew Weinstein mweinste kent edu http tamsys sourceforge net What does it look like to do qualitative research in a digital age For many qualitative researchers computers with their number crunching capacities seem anathema as research tools for uncovering textual meanings cultures as symbol systems and practices as defined n anthropology ethnomethodology cultural studies and other fields engaged n qualitative studies Yet computers have been involved in analysis of language since the 1950s and the dawn of artificial intelligence research There are now numerous programs available for qualitative researchers including NVivo Nud st MaxQDA Atlas ti HyperResearch Qualrus Transana and many others I developed TAMS Analyzer TA because there were no packages oriented towards textual analysis and
21. ase insensitive Select string Floating point numbers ode _doc ating _ Regular expression pund SST maci url GSO Cancel ound mace atting gt rural setting gt rural Ol maci yund gt cat setting gt rural si mace sund gt cat setting gt ruraliist mact Figure 8 Selecting sound gt cat After clicking OK only those records that have sound gt cat will be showing code data FRE comment sound gt cat setting rural sc sound gt cat setting gt rural s Figure 9 After selecting sound gt cat Now only two records are showing And the count indicator under the buttons shows us that 2 out of 7 total records are sound gt cat 37 To see all of the records again pick Results gt Select all That s the heart of selecting We could modify this in a whole variety of ways mostly shown in the menu displayed in Fig 6 We could switch the records that are showing and the ones not showing so only records that were not sound gt cat were visible by picking Results gt select reverse we could select additional records from the total pool of found records by picking a column and picking Results gt Select additional We could keep whittling down the records showing by doing more picking more columns and doing Results gt Select again and again Finally we could whittle down our findings by picking a column and picking Results gt Remove from selection By c
22. ata like digital audio and video but that they are not analyzed directly Instead they are linked to a log or transcript using time codes and it is the log transcript that 1s coded and analyzed TA is very smart about this and can use the time codes to leap to the important spot in the video or the other direction move the scrub bar and leap to the nearest time code in the log For my projects I ve been taping my interviews using an Olympus DM 10 digital recorder I m a real convert to this equipment The sound quality if you use the highest setting is crystal clear It s also smaller and quieter 100 quiet than the old recorders even those little micro cassette ones Once the interview is over I can connect the recorder to my computer standard USB cable and drag the file to my Mac Now this is a bit technical but I feel the need to share the only hitch with what I ve said is that the DM 10 saves the files in WMA format Microsoft Luckily there is a tool called MPlayer which is free and which if you know the right settings converts this to WAV that then can be turned into an MP3 file using iTunes Both of the last two formats are TA friendly it s just that WAV files are huge and I have no audio complaints about the MP3s which are 1 10 the size Writing field notes can be done straight into TA Document artifacts should be converted to RTF format to be useful Microsoft Word is the best way to do this though other
23. be able to generate tables that represent your searching selecting and sorting operations That is what this tutorial will concentrate on When you have a comfort with these four operations you should go on to explore data sets autosets and set operations But those topics are not for beginners 4 Searching Searching is the procedure by which you take your coded data and turn it into a nice table like database spreadsheet In this tutorial we will use the workbench for searching though each document window has an individual search tab that you can use to find information just in just that document The steps for searching for information are easy enough specify a search list which documents should the program look through specify search criteria do you want to find just one code or a series of codes or all your coded passages and press the search button This will give you a results window 1 e a database table of the passages that meet the criteria you established A Creating a search list First we have to indicate what documents our program should search through Often we have lots of documents in a project that may represent searchable aspects of our work as well as other items that we do not want to search To create a search list we use the Files Tab which contains a list of files in our project and use the buttons in the center of the workbench to move items from the file list the list of the documents in our project
24. de is not seen by TA These problems can be found by picking Check for nested from the Coding menu The moral of the story is clear run Check for pairs and Check for nested often G Saving VI TA does not automatically save document windows Save often Remember result and document windows you must save on your own Concluding comments This is only the A s of the ABC s of coding Once codes are added as described you will find them available to every file in your project TAMS provides all sorts of additional tools for working with codes There are also different types of codes to start with There are codes that describe entire documents this file is an interview there are codes that describe a section of a file this is bob talking and the codes we ve been working with which are called data codes and which identify themes See the users guide Some of the things that you might pursue through the users guide exploring the program preferences and continuing to work with your data are e How to assign colors to different codes e How to group codes together into code sets e How add comments to a particular coded passage e How to use results windows to add layers of codes to already coded passages 25 26 Analysis for Beginners with TAMS Analzyer A tutorial Matthew Weinstein mweinste kent edu http tamsys sourceforge net updated for TA 3 0 1 What you need to use t
25. es To search for a code so you can apply it just type some words into the coding field where sound gt pig is and hit return Only those codes that contain those words in the code or definition will be listed in codes If you want to expand the list of shown codes fill in a different word into the coding field and click Select The here really means more as in show more codes To see all of your codes delete any text in the coding field and hit return Code Applies the selected code to the selected text Rarely used Normally you would just select your text and double click the code name in the code list 15 New Obviously my program isn t telepathic it has to learn the codes from you to place in the code list To add a new code you type it in the little window under the button and press New See figure 2 Once a new code is added it will appear in a list under the word Codes in the above window no codes have been added to the project yet As we will discuss in the next section the general way you code text is by selecting the text you want in the right side of the window and double clicking the code from the list on the left side But this will be handled in the next section my purpose here is just to point out the anatomy of this window C Other windows There are a couple of other important windows which I m not going to talk about yet The most important of these are the windows that con
26. et Kiel dew Pru idl pig Limits Edun dog SAE ae eT Sify gt ee Two of the effects of the New button Figure 11 The Effects Of The New Button Using existing codes Once you ve taught your project a bunch of new codes applying existing codes is a breeze Simply select text and double click the code from the code list that s the list of codes on the left side of each document window We could just keep selecting oink s for instance and double click sound gt pig from the left side of the window With a second code sound gt cat the document might look like this Old Macdonald had a farm EIEIO and on his farm he had a pig EIEIO with an sound gt pig foink oink sound gt pig here and an sound gt pig oink oink sound gt pig there here an oink there an oink everywhere an sound gt pig oink oink sound gt pig 21 Old Macdonald had a farm EIEIO Old Macdonald had a farm EIEIO and on his farm he had a cat EIEIO with a sound gt cat meow meow sound gt cat here and a meow meow there here a meow there a meow everywhere a sound gt cat meow meow sound gt cat Old Macdonald had a farm EIEIO While this will be dealt with later codes can be overlapped and nested Both of the following are perfectly fine In the following example I ve also coded the place that children think animal lives with a setting code here subcoded to setting gt rural setting gt rural Old Macdonald had
27. etatags are always surrounded by braces and Y just like codes but begin with an exclamation mark Our new document begins with a critical metatag repeat You should find the project all set up in the documentation folder 27 A Context metatag The context metatag tells TA that the following list of codes are different than normal data codes These codes are used to indicate information that gives context to the data codes rather than the data codes themselves In interviews the speaker s name might be a good context code In field notes the time index or location or observer might all be context codes We are telling TAMS that when it returns information about our data we want it to include this other information who is speaking the time code for field notes or in this example the verse with the relevant data This ability to attach information to each other is one of the real strengths of TAMS but you can see that it is not exactly coding for beginners Note that the context values have to precede the data that they are coding actually if you read the user guide you can structure a document so that you can assign the context codes at different points but for unstructured documents what I have said is true the context values need to be set before the data codes Note For people who have read earlier versions of this documentation context is synonymous for the older code repeat Examples 1 Icontext spe
28. feo uc a Pare rea da are erster Je L I soan dF pig parting Suri se alia ka sendarin beiingsrursgl4 al t I send gg Grad ure aa eli u I Set inga rura peeting Surat nid 5 i soends lt al amag rurally sa eid 6 F sGend cal aig rural a P 7 2 Setting gt runs keiing gt rurs dl gdt Figure 3 A Results Window What you get is a results window This is the window that lets YOU analyze your data There are a lot of parts of a result w ndow so briefly 1 There is a panel of buttons that you can configure and by default includes buttons for taking you back to your project window Workbench back to your document windows find record and for refreshing this window if one of your documents changes among other functions 2 Then is a status panel which gives you information about your results It tells you how many records are showing a check box which tells you whether any of your documents have changed which means you should hit the refresh button and a description of the data you re seeing In this case it tells us that we have done an unlimited search At the far right side is a little button that is used for doing Select near analysis 1 e give me all examples of people talking about pigs near within 5 lines or any definition of near you develop people talking about goats 3 Below the status panel is a large pane that shows us the data of the selected record To select a different record click on a different row of
29. gt cat etc I have to click the Exact box under the search Otherwise TAMS assumes I want the whole family Finally and this is almost a beyond beginner item If I want passages that have two codes 1 e the intersection of coded passages I would need to use a plus sign between the codes In other words if I want to see the part of my document that is sound gt pig and setting gt rural at the same time I would type sound gt pig setting gt rural into the search field If you feel comfortable with all of this you can learn a few other tricks to searching by clicking the search tab on a document window and looking at the Help box 5 Sorting Finding information is the first part of analysis We re half done with the topic of finding information but before continuing I want to offer as an interlude some notes on arranging information This is the art of sorting and TAMS can do very complex sorts While beyond this tutorial TAMS can also remember sort orders and use them in macros that are very powerful for finding key information across searches I ve already noted that a simple way to sort s to click on a column in a results window and hit the sort button The problem with this method of sorting is you have no control over the type of sorting that the program is doing The more powerful sorting tools are under the Results gt Sort up and Results gt Sort down menus These allow you to control the direction of the
30. gths and weaknesses almost instantly So what you learn here will be different at least slightly from what you would learn with other software but there are commonalities as well This first chapter introduces how TA works at the most conceptual level 1 e how it links to the qualitative research process This chapter will not have any screen shots and few tricks It is about how to think about using qualitative research tools in the course of a research cycle To tell this story I have flattened out the research process But I think this pattern and growth in the research cycle will be recognizable to those that have done qualitative studies with TA and probably similar packages I Phase 1 In the field and on the computer Most qualitative data sets are defined by 3 types of objects interviews observations and artifacts While these can t be considered comprehensive surveys just as a start they still dominate those fields engaged in qualitative research projects Obviously depending on the question being asked the balance of these types of objects will vary For the studies I ve recently become involved in interviews have often provided the bulk of data Digitization starts in the field Field notes are often typed straight into laptops these days Digital video is increasingly what is analyzed rather than the observation notes themselves TA is a text oriented package This doesn t mean that you can t use other forms of d
31. he workbench or project window ties your files together The purpose of the workbench is threefold This is the window you use to define your codes It is also the window where you can put together multi file searches Finally it is the window that allows you to save and recall your whole project a topic we wont explore in this introduction see the documents for a brief discussion of saving projects The Files tab has opetions for add ing files to projects opening exist ing projects and assembling multifile searches The info tab is used for setting information for multi user projects Use the search tab for mining the information you ve coded for autocoding and otherwise exploring your data Define codes allows you to create define and set colors for the codes you use in your project A Pe flankFrej_ et pr Fine files Search Define codes Code se Prodi name Cas cri pce Code sets provides a means to group codes together into larger themes Career Gabe tyes BSL SaL Port Database name User Initiate Pareo Largest file size J Mag onic Report Ta project info Manage users Figure 1 The Workbench We will begin by clicking on the file tab so we can create a new document in our project The workbench will then look like this 10 The left hand panel The right hand panel and and buttons are for buttons are for managing the managing the files in Search list i e the list of files your project i
32. his documentation At this point you should have read through document called Coding for Beginners If you have hopefully you know the difference between workbench windows and document windows you know that I use project windows and workbench windows interchangeably and can now create documents enter data create data codes and apply those codes to sections of your document We ll be using a modified version of the Old McD from Coding for Beginners for this analysis which is merely to introduce you to the basic procedures for analysis context verse verse 1 verse setting gt rural Old Macdonald had a farm EIEIO and on his farm he had a pig EIEIO with an sound gt pig oink oink sound gt pig here and an sound gt pig oink oink sound gt pig there here an oink there an oink everywhere an sound gt pig oink oink sound gt pig Old Macdonald had a farm EIEIO setting gt rural verse 2 verse setting gt rural Old Macdonald had a farm EIEIO and on his farm he had a cat EIEIO with a sound gt cat meow meow sound gt cat here and a meow meow there here a meow there a meow everywhere a sound gt cat meow meow sound gt cat Old Macdonald had a farm EIEIO setting gt rural This version of our old data introduces some very not for beginners ideas The first of which is the metatag A metatag is an instruction to the tams program It s the way the researcher specifies how TAMS should treat the data M
33. ind that I want to relocate the beginning or end tag I ll read the next paragraph and realize that should be included in the coded passage Start by selecting the code One way to do that is to click in the middle of the tag and pick Find current code from the menu That will select the whole tag Now you can use the mouse to drag and drop the end tag to a new location You may also want to find the other end of a code pair By a code pair I mean the front code that is the one that looks like a and the back code the one with the slash a Just click in the middle of one end and pick Find paired code from the Coding menu 25 Sometimes you may want to just move through the document code by code This is easily done with repeated use of Find current code and Find next code Deleting code pairs is also something that TA makes easy If you have just inserted the code by double clicking or using new you can pick undo from the edit menu though for new codes this will not do anything to the code list only to the document Alternatively you can click on one of the tags 1 e anywhere inside the braces and pick Delete code pair from the Coding menu Finally all of the codes in a selection of text can be removed by 1 selecting the portion of text you want stripped and 2 picking Remove codes from selection Problems coding There are a number of problems with coding that can crop up and T
34. is that it is a sandbox you try things see how they look and then try different things The dialogue wont go away until you click exit So it s really designed to allow you to experiment til you have the data the way you want for analysis or for reporting It is non destructive to your data 38 Just to illustrate the most basic features of this report I m going to generate a table from Old Macdonald project of all the pig sounds So I select from my results window the pig sounds by selecting the _code column picking Results gt Select and filling in pig and hitting return Now pick Reports gt Data comparison table Across the top we ll put the codes that we have selected so make sure the pull down menu in the area pointed to in box 2 in Figure 10 says Codes We ll break it down by file along the vertical axis so pick _doc from the pull down menu indicated by box 3 Now we have to pick what we want to see in our table In this example we want to see the actual data for that code so select the _data bar from the table pointed to by box 4 in figure 10 Now just click Show report This is what you should see in your default web browser _doc compared against Codes Figure 11 Results from data comparison table In this simple foolish example we see that in oldMacD rtf the sound gt pig code was used 3 times all with the same data oink oink As a second example I m going to look
35. ith an oink oink here and an oink oink there here an oink there an oink everywhere an oink oink Old Macdonald had a farm EIEIO HEICOCOC 2 0 Code to toolbar Def Ruler see Text area where you type in your ie ie 3 data or import it from another sound gt pig word processor TA can use both a rtf and text documents Codes sound gt pig Coding side of the document window where new codes are added and old codes are listed Figure 4 The Document Window And Its Parts Edit mode The buttons are worth a little explanation The ruler button toggles a ruler in the Text Area the right hand side the other five buttons deal with codes 14 Code to toolbar Def C Ruler Select Code 1 New sound gt pig Type in a new code Codes here and press the New button Existing codes will be listed in the panel here Figure 5 Adding A New Code Using The Button Panel Code to toolbar This button will take whatever code is selected in the code list and make a toolbar button for it Makes it easy to access codes you use a lot Def This button will pop up a definition of the code in a window to remind you what your codes meant when you first used them Select One of the most powerful and almost hidden features of TAMS Analyzer is the hot code list This feature is most useful when you ve created dozens or even hundreds of cod
36. ld be done with any number of tools Before I wrote TAMS Analyzer for instance I used a word processor for coding and a small program still available at the TAMS website to pull out the information I needed then I used programs like Excel to do the actual analysis of that data To mark up text you surround the part of the text you are interested in by tags which have in them a code Together they look like this mycode To indicate the end of the section of text you re interested in you put another tag with a in front of the code name mycode So in our animal example we would type into the text sound gt pig y before the words oink oink and sound gt pig after those immortal words oink oink Now that text has been coded It would look like sound gt pig yoink oink sound gt pig This means in TAMS that oink oink is an example of sound subtype pig You could do it with any wordprocessor What is TAMS Analyzer aka TA TAMS Analyzer was my attempt after using TAMS for a while to create a more complete application for coding searching for codes and IV recoding going back through and adding levels to the codes TA is still not the whole megilla it doesn t have graphing for instance For many projects you will still want to save the results and use Excel or other programs Panorama is a wonderful database for this sort of analysis to do the more refined counting
37. mple of finding how many sound gt cat codes there are out of the total number of coded passages To start with do an unlimited search Go to your workbench and press the clear button and then the search button You should now have a result window with one line for every coded passage in the document one line per code We can see that we coded 7 passages not counting our repeat codes by looking right under the buttons on the results window Workbench Find re 717 Unlimite jonald had a farm EIEIO a pig EIEIO k nink coiingdsnios here Figure 6 We found 7 coded passages and all 7 are showing Let s find the sound gt cat codes Pick the _code column by clicking on its header Then pick the Results gt Select menu option Font Coding Project Me elatags Reports Wine l 2 IOOF EAG Hide Lass E El show Comment E Set field to browse FD meee Eu Play mel Ua riid ig Zack E eT ET Farel Zero memory Refresh Find record in deinen Sort options Set i h cie ost TATI E dat soend gt gig nid Warked records b fiend gag oki Did fiend gt eg TEET p b ok Hring pid E ziad Seleci if HE wd mendeg Select additional nidk Tid ira chon oidh Select reverse Select all THA Select ney Switch data view Salat frase ce Figure 7 Picking Select 36 This provides a dialogue box into which you can type the phrase you are looking for sound gt cat sount select sound gt cat F gt Search c
38. nterviews that are examined when you want field notes transcripts etc to analyze the pattern of codes in your text ana BB Childress songs prj Search Geline code Crate sal TG nt Arts all The init button is used to designate which file in the left hand panel is the first file searched Seth modo F These buttons are used to add files to and These are used to share files remove them from the in multi person projects search list and Figure 2 The Files Tab Document Windows To make a document window you need to click the new button on the project window The new button is right over the Files list view on the right side of the workbench window under the dividing line 11 New Open sets Holder Files Figure 2 The New Button On The Workbench Files Tab Makes Blank Documents Windows Click the new button and you will be prompted for a name Type MacDonaldsFarm Children s songs xtprj nfo Data file name MacDonaldsFarm Add gt Addall gt gt Figure 3 Naming Data Files Once you click Ok you will see an almost blank document window Starting with TAMS Analyzer 3 0 the window actually starts with a little bit of helpful text at the top of the window Leave it there and start your data below the text in braces The document window is a sort of specialized word processor or text editor designed for coding
39. ode is Now Selected for Searching or Selecting TAMS can sort alphabetically numerically either by integer or real floating point comparison by date though you must pick the format first by picking it from the Results gt Sort options gt Date format menu item first it s best if there s only one date format in a project and by code This latter type of sort is really not part of a beginners tutorial but just so that you know if I set the code level to 1 by picking the Results gt Sort options gt Code level dialogue sound gt cat and sound gt dog will be taken to be identical for the purposes of sorting since they match at the FIRST this s the meaning of code level 1 level of codes Picking code level 2 means both level 1 and 2 must match Ok that s for another lesson Now you can organize your data as well as find it 6 Selecting The third type of operation you need to know about is selecting Selecting can be thought of as finding information inside of a search It s a way of searching searches or better 35 stated searching results files Remember that a results window is a fairly complete database and a good database should be searchable The terminology is confusing because selecting also refers to the way that we pick things on the screen with our mouse select column _code select the first field etc Ill try to use pick instead of select for the latter type of operation Let s take as an exa
40. ombining these and using different columns your _code column your repeat columns and your _ data column you can find many patterns inside your data 7 Comparing data When you have searched sorted and selected your data you will want to see it TAMS Analyzer lets you browse your data one row at a time and see the results toggle whether the tags are showing etc But it s important often to see all of your data in one table side by side so you can see how different interviewees responded to a question or address topic code X TAMS has a very powerful tool that generates html tables that you can copy and paste into your word process called the Data Comparison table After you have the selection of data you want to examine pick Reports gt Data comparison table This is the dialogue you will be confronted with 1 Switches the m rows and col Fe Wien use cave nude si 2 By default sets the ee criteria for the col Fu umns of the table 3 By default sets the ae eee criteria for the rows of the table 4 B r team Ei 5 Controls the justifi 4 Sets the criteria ze cation for the differ for what is filled in ent parts of the table the inside of the table Figure 10 Data comparison dialogue While this is a very complex dialogue it is also very powerful It can provide counts and subcounts as well as show you the data or other information codes comments that you are interested in My metaphor for this dialogue
41. qualitative research for Macintosh OS X when I desperately needed one in 2002 I have continued to extend and develop the program over the years Note that I have not used any other qualitative packages and as a result TA is a bit sui generis However I did use computers for qualitative research while working on my dissertation in the early 1990s Weinstein 1998 I used a programmable word processor called Nisus an outliner called More for analysis and Panorama a very flexible database good at manipulating large chunks of text quickly Mixing and matching programs was important for me then In 2002 none of the programs I had used had been updated for the new Macintosh operating system TA was my answer In many ways TA s design with open in text codes reflects the heterogeneous way I dealt with data early on The codes are naked ugly visible for anyone to see and any program to use It s a virtue and a handicap Somewhere I read that unlike quantitative software every qualitative package is different Skills developed on one package don t easily transfer to others The language different software programs use to describe their operations varies as well Qualitative research 1s rather non paradigmatic in Kuhn s sense of the word we don t share a common language problem model solution model world view That said when I met with the creator of MaxQDA recently we were able to compare features and assess our software s various stren
42. re here an oink there an oink everywhere an oink oink Old Macdonald had a farm EIEIO I Hl We would like to mark oink oink as the sound a pig makes Now we may have a lot of codes and collect a lot of different information including a lot of peripheral information about what children think animals are so we need to design our code system carefully so we ll be able to keep track of all the information we ll have coming in For our study whenever we have an example of a sound we ll mark it with the code sound and then a subcode for the type of animal it is So we would want to mark oink oink with and I m using TAMS sytax here if you use a different system this would be different with sound gt pig If we get information about what children know about pigs we will mark it dea gt pig How we do that will be described later but it could be done a lot of ways sticky s hanging off the side with idea gt pig on it for instance Computer coding usually involves selecting the text and somehow picking the code from a list What is TAMS TAMS which stands for text analysis markup system is simply a way of indicating in texts what the codes you re using are It looks a lot like html and xml which are languages used for making web pages and I certainly was influenced by those ways of marking up text The idea was to make a system that was easy to use easy to see and flexible enough that it cou
43. sort Should A or Z be first If A is first you are sorting up if Z is first you are sorting down and allow you to nest sorts e g first sort A Z on one column and then sort A Z on a different column That way you can sort by file name and then sort by the codes the result will be all of the codes sorted by file name 34 Alternatively you may put three or four documents in your search list and then want to see how many documents use each code So you want to sort by the codes and then sort within each code by document Specifically you would click on the column with the header _code and then pick Results gt Sort Up gt alpha 1 e alphabetically Then you would click on the header of the column labeled _doc and pick Results gt Sort up gt alpha within The word within means keep the first column sorted appropriately but rearrange items that match in the first column according to the currently selected column You could sort by any number of columns within each other If you want to start over just sort NOT within the others Just pick Results gt Sort up gt alpha or whatever item is appropriate eade _data _comrr setting gt rural setting gt rural l sound gt pig setting gt ruralys sound gt pig setting gt rural s sound gt pia setting gt ruralys setting gt rural setting gt rural Ol sound gt cat setting gt ruralys sound gt cat setting gt ruraly si Figure 5 Column _c
44. tain your results by which I mean the results of searches for relevant data This will be discussed in Analysis for Beginners with TAMS Analyzer a document also included with the program V How do I code in TA A Adding codes with the project window We ve seen one way to add new codes select some text fill in a code into the code field and then click the new button Sometimes however a researcher comes to a project with codes already in mind and needs to enter them even before transcribing an interview and applying the codes as you find them One way to add such a priori codes is through the define codes tab on the Workbench To find this move your workbench to the front and click on the tab called Define codes The workbench should now look as follows 16 ee e F Children s songs xtprj Info Files Search jefine codes Code sets Memo Mail Codes Name Color mn Mi Use inherited color M Active History Delete colors Delete Save Clear remem O T ere Figure 6 Creating New Codes The name of our first code will be sound gt goat We ll enter that next to Name In the big box underneath the name box enter the definition Marks text showing what children think goat sounds should be Name sound gt goat Color gum Use inherited color M Active Marks text showing what children think goat sounds should be Figure 7 Filling In New Codes
45. the table below This is the browser pane 4 Finally occupying the whole bottom of the window there is the table of records one row represents one passage of text that met the criteria we asked for A simple unlimited search returns a row for every coded passage in our data If a section of text has two codes codel code2 This is my text code2 codel for instance that text will appear in two different rows in the table of records one for codel and one for code2 yes this might mean that there is redundancy A non simple search removes the redundancy and returns one row for every stretch of text surrounded by codes no how many codes are embedded in that text but you lose information that way By clicking on one of the rows you can use your up and down arrow to browse through your data examining the passages in the browser pane If you want to look at one row of the data in its original context double click the row Finally it is worth examining the columns given to you in the table of records The first column is a simple numbering of the shown records Next will appear all of your universal and repeat values Here we can see which verse each coded passage came from If we had an interview we could see who was speaking Next comes the code of the passage followed by the data followed by comments we ve attached to the codes another non beginner function the name of the coder if you are using multiple coders the name
46. to identify which of the codes I created were related to teaching science for citizenship rather than as a way to make more scientists These codes were in various code families but all related to issues like empowerment equity health and environment I created a set of those codes and was able to determine that there was no particular gender sex bias to embracing those themes I had not done gender analysis of this data prior to this but it took about a minute to tell TA what the sex gender of each of my interviewees was this was done using an if command in TA if the interview is with Bob sex is male if the interview is with Sally sex is female Then having created the code set I ran a search of all my coded passages about 10 seconds and then asked for a concept map a graph of the relation of sex to social issues I also looked at how social issues broke down by each of my interviewees The point is that I can ask questions about these large ad hoc cross familial categories by getting either A maps or counts of codes or other variables context codes I create against these code Sets This has described a pretty typical research cycle using TA There is one more part of this that deserves a comment or two and that is how do you get images out of TA that communicate what you have discovered IV From analysis to print Having done the analysis reanalysis and code set analysis you
47. tting any limits on what data is returned we want it all This will demonstrate how TAMS turns raw stuff interviews into tables To do an unlimited search make sure that the field marked Search is empty that the pop up menu underneath the check boxes says Simple and leave the raw box checked and the other two unchecked and press the button marked Search 30 ae Fl OldMacD xtprj l Info Result file name DOO Oe Mail Z Temporary Cancel Ok Named search Clear Coder IDs P Add current code set Ed Raw _ Empty _ Exact I Simple Figure 2 TAMS Prompting for a Result File Name TAMS will now prompt you for a name for your search Just leave the field blank don t touch the Temporary check box and hit the Ok button This will automatically create a temporary result file 1 e the next time you run TAMS it will throw out these results This is typically what you want Most results are exploratory and are not intended for permanent storage If you do want to save the results permanently you can use the File gt Save as option to give the file a real name and making sure that Temporary is not checked That way you can keep the results file Ana Ei temoResult 21008008 BABL 1109 6E3 00003 MOF IAC wis a w El PM EEE ty Cun mine Suse Borkcanch Find recors Fe s k Jari Select ul Sabet rir IEIET n I nti neg el eu ned gt ig oink oink
48. u should also be able to designate some codes as context codes to identify contextual information From here some simple additional things to learn include personalizing the toolbar on results windows adding and viewing comments in results windows and exporting data to the clipboard so that it can printed in Microsoft Word Appleworks or the word processor of your choice An even more advanced step would include learning about reanalysis 1 e transforming your data files based on your results files This includes recoding your files actually changing the code of a given passage and adding codes around given passages Finally there is a lot to be explored in terms of reporting including simple reports counts of codes and complex reports data summaries and dot graphs These are all 42 described in the user guide Have fun exploring this complex but powerful software program 43
49. ve code B turns out not to be the same question as how many passages coded with B also have code A Why The first time we would get the passage marked as orange as the result of the first and second search R school gt trouble Well my highschool was known as a trouble school aspirations We weren t going anywhere aspirations violence There were a lot of fights violence and truancy kids uhm wandering around truancy and aspirations most of us worked in factories on the city s east side aspirations gratification gt delayed Most of us partied rather than worked gratification gt delayed school gt trouble The second time we would get the following two orange passages returned separately R school gt trouble Well my highschool was known as a trouble school aspirations We weren t going anywhere aspirations violence There were a lot of fights violence and truancy kids uhm wandering around truancy and aspirations most of us worked in factories on the city s east side aspirations gratification gt delayed Most of us partied rather than worked gratification gt delayed school gt trouble If you want a count of intersections of the two codes rather than a count of how many of coded passage X also has some of Y you should do a search from the workbench with the sign aspirations school gt trouble 9 Concluding remarks At this point you have some idea of how to search sort and select Yo
50. ve data in a digital age For TA it means keeping track of lots of themes and sub themes and uber themes and cross themes As with many digital issues the payoff comes with scale If you re analyzing one short interview Post Its are probably the way to go If you re dealing with 3 100 interviews Post Its are out of the question and it s time to use TA or one of its cousins Now for the details how do you go about creating codes importing or creating documents linking media files to logs and search for coded information The basics will be described in the rest of the For Beginners series the details expanded upon in the user manual and release notes Creating a project Generating initial themes ranscribing Importing documents New project Define codes New amp import document Transcriber On the fly code creation Define codes On the fly code creation Search for codes Selection of codes in searches Near selection in searches String searches Data comparison table Summary reports Coding for Beginners with TAMS Analyzer A tutorial Matthew Weinstein mweinste kent edu http tamsys sourceforge net updated for TA 3 0 This is not comprehensive documentation of the TAMS Analyzer It is more of a tutorial or actually the necessary nuts and bolts to get going on a qualitative research project From this documentation users can read the documentation for the program and the coding system I
51. will want to share your findings with the world i e the other 5 people as fascinated by your topic as you TA has a wide variety of reports including code counts cross counts of codes how many code X and code Y apply to the same data counts or concept maps graphs of the relation ships among codes context codes and code sets Most of the table style graphs can be saved as a tab delimited text files perfect for Excel or Word the code set report is generated as an html file which can be copied and pasted right into Excel to generate sums averages etc The concept map style graphs are generated through a program called Graphviz app which can export graphs in many graphic formats tiff jpg etc To be honest most of the graphs generated by TA I use for analysis not for presentation I do count on TA to find the exact quotes I want to use in my papers and TA allows me to copy the quotes without all of the coding information which my readers don t necessarily want to see When I have needed a graphic a pie chart of code counts for instance I ve saved the counts to a text file and used Excel to do the actual drawing rather than TA Remember that one of the inclinations behind TA s design was to make sure its data was available for other programs to do some of the heavy lifting V Conclusion My idea behind this short piece was been to show how a researcher might use software to make sense of qualitati
52. y these searches are used for analysis but at this point in the process they are used to remind yourself of the meaning of a code The search returns a window which lets you page through each instance of the code applied to a passage of text Reading the passages you ve coded a particular way allows you to inductively recall the meaning of the code For complex themes more than the definition this gives you clues as to tone topic theme the code was intended to identify Once you have coded all of your documents and identified the major themes it s time to deepen those themes Given a theme like negative behavior TA actually doesn t allow spaces in code names so it would probably be negativeBehvior or negative_behavior you would want to further classify each passage as to the type of negative behavior This part of the process I call reanalysis Reanalysis involves looking at the instances of a particlar code to alter your original coding scheme To do a reanalysis you do a search for the code you are interested in breaking down and then lock out most of my editing capabilities by throwing the software in reanalysis mode hint it s on the project menu This will keep you from doing anything to your source files that would make them out of synch with your search results Next you ll sort through the results of that search and try to identify the key types of negativeBehavior If one type is name calling then you will
Download Pdf Manuals
Related Search