Intro to NVivo 8 workbook. Helen Marshall

Transcription

1 Intro to NVivo 8 workbook Helen Marshall s in this workbook? The two workshops for which this workbook is written are for researchers wanting a first meeting with QSR NVivo8. You may take only the first workshop, or both workshops according to your needs. Their aim is to give you an overview of relevant software tools for some of the common research tasks. The assumption is that you have some understanding of those tasks. Each workshop is divided into topics, exploring at a basic level some of the tasks a qualitative researcher faces. This workbook has basic instructions for carrying out these tasks and advice on going on to learn more or start your own project. The advice is in the list of resources at the end of each topic. They usually include the relevant headings in NVivo help. They also have a reference to the NVivo 8 tutorials free online at http.//wwwqsrinternational.fileburst.com-document-nvivo8- Teach_Yourself_NVivo_8_Tutorials.pdf. These tutorials contain useful methodological discussion as well as practical advice. I reference the chapter of Lyn Richards Handling Qualitative Data, (Sage, London, 2005) for which those tutorials were written plus the relevant section of Pat Bazeley s Qualitative Data Analysis with NVIVO (Sage London 2007). While the books both refer to NVivo7, they are still useful for 8. Pat Bazeley s website contains some updates for 8, and Sage will be publishing a new edition of Handling Qualitative Data later this year. Workshop Topic DAY 1 Before we begin- exploring an NVIVO project. 1 Setting up a project. 2 creating documents in NVIVO and bringing in material from outside 3 transcribing or making notes from media sources in NVIVO, 4 memos annotations and see also links 5 Organising sources using sets 6 Coding, free nodes, tree nodes, coding stripes. 7 Speeding up topic coding using autocode. 8 Creating cases and attributes DAY 2 9 Using find, 10 text search queries, 11 word frequency queries 12 coding queries, 13 Running a matrix and showing matrix results visually, 14 Using relationships 15 Running Reports 16 Models RMIT University Page 1 of 25

2 Before we begin: An overview of NVivo OPENING AN EXISTING PROJECT Open NVivo click on volunteers. Navigation Pane List View Detail View The The screenshot above shows you that an NVIVO project has several elements. In Volunteers, practice moving between navigation pane, list view and detail view. See how menus become active as you move around a project. Notice three key elements of an NVivo project in the Navigation pane. Sources are the data for a project. Look at what s in volunteers. Nodes are the containers for snippets of data. You can highlight a bit of a source or all of a source and tell NVivo to put it (actually a reference to it) in a box. Look at the various kinds of nodes (tree, free and case) in Volunteers, and open some up so see what s in them. Queries are the tools you use to interrogate the data. Look at the queries in the list pane of Volunteers to see what has been done, and then find the results folder to see what the outcomes look like. RMIT University Page 2 of 25

3 Topic1: Setting up a project OPENING A NEW PROJECT Open NVivo - File menu - New Project. Get into your empty project, and take charge. Think about the options you want for access. Get comfortable with the navigation process and customize appearance. Think about how you want to save and back up your precious work. CUSTOMIZING ACCESS CUSTOMIZING APPEARANCE SAVING COPYING (For backing up to a different location, e.g. a USB) File - Project properties Tools Customize - Options File - save. Close Project File - Copy Project Resources for topic 1 Within NVIVO 8: Go to Help and the topic on Setting Project Properties. Teach Yourself NVivo 8: Tutorial 1 Setting up Your Project Richards, 2005: chapter 1 Bazeley, 2007: chapters 1, 2 RMIT University Page 3 of 25

4 Topic 2: Creating documents (in NVIVO and bringing in material from outside) Making documents in NVivo is easy. A date-stamped project diary is always a good idea for capturing ideas on the run and noting what you did about them (very handy for the methodology chapter!) CREATING DOCUMENTS DATE STAMPING DELETING DOCUMENTS ALTERING DOCUMENT PROPERTIES Sources (select folder) - New (In the Source) Format - Insert date/time Select source Edit - Delete (List view) Select Sources RMB- properties You will want to bring in data from outside your project too. You can bring in formatted Word documents, and import media (.mp3,.wma,.wav,.mpg,.mpeg,.mpe,.avi,.mov,.qt,.mp4) and PDFs. (PDFs are saved as text. Format will vary and you may encounter problems). IMPORTING DOCUMENTS OR MEDIA (List view) Sources - Internals (select subfolder if you wish) - Project - import internals. When you are importing sources, think about whether you wish to have NVivo create a case node for each source. Do this if you are going to compare one source with another (e.g. one individual with another.) If you are comparing grouped sources (for example, you will compare data from the interviews, focus groups and observations in organisation A with data from organisation B) wait till you have imported all the sources, then create case nodes for A and code all the relevant sources at it. Repeat for B. Resources for topic 2 Within NVIVO 8: Go to Help and the topics on creating documents and importing documents. Teach Yourself NVivo 8: Tutorial 2 Creating and Importing Sources. Richards, 2005: Chapter 2 covers bringing in data. Bazeley, 2007 Chapter 3. Also look at her website for handy notes on using word for formatting documents: RMIT University Page 4 of 25

5 Topic 3: Transcribing or making notes from media sources in NVIVO Of course not all your data comes ready transcribed. You can use NVivo for transcribing, or more likely for partial transcripts, with constant access to your media files within the project. (Note however that lots of media files make for a large project.) TRANSCRIBING STOPPING AND STARTING USING FUNCTION KEYS RATHER THAN MENUS ADDING A TRANSCRIPT ENTRY Open your source (from media toolbar select) Play mode - transcribe - play speed - slow F7, (Play) F8 (Stop) F9 (skip back) In an audio or video source, click in the transcript existing entry. To add an entry below it. Format Insert-Row. To add an entry for a selected time span click and drag to select a time span on the media timeline then Format Insert-Row. Resources for topic 3 Within NVIVO 8: Teach Yourself NVivo 8: Tutorial 2 Creating and Importing Sources Bazeley covers transcription in the additional notes to her book. You can get these from her website. RMIT University Page 5 of 25

6 Topic 4: Memos, annotations and see also links You often want to add notes of various kinds to your sources even before you start coding. For example, you might want to add notes about the context of an interview to the transcript. You can do this with a memo. Memos are documents that you can create and then link to another document, or can create already linked to a document. Note that you can only have one memo per source. LINKING A MEMO TO A DOCUMENT Select the source : RMB or Links menu - New memo link Memos are not the only kind of note you may want to link to a document. For example, you might want to create a sort of footnote - the equivalent of writing a reminder about the text in the margin of a transcript. You do this with annotations. (They can be really useful if you want to explain why you have coded this portion of the document under one heading rather than another) ANNOTATING highlight relevant text- RMB or Links menuannotation new annotation As you do your preliminary thinking about your sources, you may find you want to link the contents of two sources, or to make a link between two segments of one source. For example, you might want to link a bit of an interview to your notes about a theory you encountered in the lit review. On paper, you d write see also XXX. Or you might want to link the two segments of an interview where your participant seems to contradict herself. On paper, you d draw arrows or write see also page In NVivo, you use see also links. LINKING TO A DOCUMENT LINKING TO PART OF A DOCUMENT highlight relevant text- RMB or Links menu- see also link new see also link - select document Work backwards! Open the source or node you want to link highlight RMB or Edit menu-copy. Open and highlight the portion of the item you want the link to come from. RMB or Edit menu - Paste As See Also Link. Resources for topic 4 Within NVIVO 8: In help there are separate topics on Annotations, Memo Links and See Also Links. Teach Yourself NVivo 8: Tutorial 4 Richards, 2005: Chapter 4 covers annotating, memos and linking data. Bazeley, 2007: Memoing etc is covered in various places especially chapter 4. RMIT University Page 6 of 25

7 Topic 5: Organising sources using sets As your sources grow, you will need to manage them- for example to collect together the stuff you have not yet read! Sets are a (perhaps temporary) way of grouping your sources (and later nodes) member. Items in a set are references or 'shortcuts' to the original file so you can delete an item from a set without removing it from your project. A set can include any number of sources or nodes and a source or node can belong in multiple sets (which make sets different from folders- use folders for permanent organization.) MAKING SETS Sets - New ADDING TO SETS Sources - add to set DELETING A SOURCE FROM A SET Open a Set - Click the required items. Click and drag to select multiple consecutive items or hold down the CTRL key to select non-consecutive items- Edit Delete Yes Or ( Right-click the required item(s) in List View Click the Delete option) Resources for Topic 5 Within NVIVO 8: Help, Sets. Teach Yourself NVivo 8: Tutorial 3 Managing data: Cases, Attributes and Sets Richards, 2005: Chapter 3 for advice designing the use of attribute data, and on the ways of grouping and focusing on data. Bazeley, 2007: for an overview of managing data using cases, sets and folders pp111-4 on sets. RMIT University Page 7 of 25

8 Topic 6: Coding Coding is basically just indexing, and all qualitative researchers do some of this, even if they never commit their index to paper, let alone to computer. You will make your own decisions about what to code (all your sources or only some). Some projects rely on inductively coding up from data others on deductively coding down from theory. Probably you will sometimes lump material into broad categories and sometimes split it into fine ones. You will have both topic codes that signal this is where in my data and reading I find theme X and analytic codes that signal this is where in my data and reading I find stuff related to concept or theory Y. You can code from your sources, or you can code on from an existing node. The software doesn t restrain you. In NVivo, two project items are always involved in coding the source whose content is being coded, and the node at which you place a reference to the relevant material. Nodes can be created in any of four folders, Free Nodes (stand alone boxes free of organization), Tree Nodes (stored in hierarchical catalogues), Cases or Relationships (both covered later). It s usually wise to use the properties box to define your nodes. TO MAKE AN EMPTY FREE NODE Nodes - free RMB or New Menu New Free node TO CODE TEXT INTO YOUR FREE NODE TO CODE DIRECT FROM AUDIO OR VIDEO INTO YOUR FREE NODE TO CODE NOTES OR TRANSCRIPT FROM AUDIO OR VIDEO INTO YOUR FREE NODE TO MAKE A FREE NODE AS YOU CODE TO CODE ON FROM A NODE Open Source - Select text (take it slowly) use RMB, Code menu, the coding bar or drag & drop to code the text into an existing node. (For drag n drop, you might want to set the view so that your detail view is to the left). Use the play controls on the Media toolbar to locate the portion of the audio or video you want to code. You can also drag the media playhead to the required position on the timeline. Click and drag the timeline to select the required timespan. RMB or Code menu code at existing node. Select text use RMB, Code menu, or coding bar but you can t drag & drop. Open Source - Select text - use RMB, Code menu, or coding bar. Coding In Vivo names the node with the selected text (only possible for documents.) For media, select the portion of the audio you want to code. RMB or Code menu code at new node. Open the node, - Select the text and code RMIT University Page 8 of 25

9 You make the decisions about how much to code (though as a rule of thumb, going for a paragraph rather than selecting only a few words is probably sensible). Should you want to get back from the node to see the words in context, you can do that easily. TO SEE THE CODED ITEM IN ITS CONTEXT In the node, hover over the link to the source and Left MB or select the text RMB As your categories proliferate, you will probably see connections between some of them. NVivo allows you to create hierarchies of nodes. They are called tree nodes, and NVIVO also uses the language of parent nodes (like images of volunteering ) having children (like adventurous community minded and energetic). A tree structure can help you see (and show others) how your ideas are shaping. You can create tree structures before you do any coding or as you code, or you can move free nodes around as you begin to see connections between your stand alone boxes. Warning: hierarchies with more than three levels may not be useful. (Do you really want to go looking for the great-great-grandchildren nodes?). Coding into tree nodes is exactly the same as coding into free nodes. Think about whether you want your parent nodes to be empty, or to have in them all the stuff that is also sorted into child nodes. TO MAKE A NEW TREE NODE Nodes - tree RMB or New Menu -New Tree node. Make child nodes from your parent node. TO MAKE A TREE BY MOVING NODES AROUND Nodes, then in list view Cut or copy and paste nodes into a structure that organizes your ideas MERGING NODES Select the node/s whose content you want to merge RMB or Edit menu copy (or cut). Click the target into which you want the content to be merged. RMB or Edit Menu mergeselected node or new node. If you copied, you might then want to delete the original node. RMIT University Page 9 of 25

10 You will quickly create a lot of nodes. You can open any node and view its contents, which will give you an idea of what sources have been coded. You may also want to see how a particular source has been coded. You can review your coding using coding stripes. The density stripe is a quick guide to how much work you have done on a source. (If you aren t sure whether you ve coded the last paragraphs yet, look for the dark line!). Nvivo8 coding stripes also enable you to explore demographic attributes. For example, you can display a stripe to show everything coded within a node from cases to which the attribute value 'male' has been assigned. TO VIEW YOUR CODING using stripes TO VIEW YOUR CODING highlighted in the text In a source use View Menu - coding stripes or the coding stripes button In a node (to see attribute coding) view coding stripes or use button - selected items Open desired node or source use View Menu - Highlight. Coding for the selected nodes is highlighted in yellow. To specify that you want coding automatically highlighted when you open a source or node Tools- options-display- highlight coding. Methodology note: Here s Pat Bazeley s good advice on trees: (2007, p105) Organise them round conceptual relations. Put like things in the tree- E.g. all these free nodes are really about images people have of volunteers. Use separate nodes for different elements of the text e.g. people, events, actions etc see pp106-7 on this Each node should have only one concept in it ( adventurous is separate from passionate ) Each concept should only be in one tree ( adventurous; is an image, it s not one of the benefits of volunteering that could be having adventures ) Expect to code one passage at several nodes Go for a flexible system- avoid long hierarchies and too much detail. Resources for Topic 6 Within NVIVO 8: go to Help topics coding, Merging Nodes RMIT University Page 10 of 25

11 Teach Yourself NVivo 8: Tutorial 5 Nodes and Coding, Richards, 2005: Chapter 5 on coding Bazeley, 2007 Chapters 4 and 5 (especially pp on coding) RMIT University Page 11 of 25

12 Topic 7: Speeding up Coding using autocode Nothing does your analysis for you, but NVivo will do some boring indexing automatically if you have formatted your text sources. Autocoding using headers can enable fast identification and coding of topics giving you more time to think about the analytic coding. Autocoding requires that your text documents be formatted using heading levels. (For more on formatting see resources, especially Bazeley). A node is created for each paragraph that is formatted as the selected heading level and the text under the heading is coded at the node. Text searching to find material is useful in limited circumstances e.g. when you want to identify topics like use of a specific medicine. TO AUTO CODE FORMATTED DOCUMENTS In list view, select the sources - RMB Code Menu - Auto Code - Headings - Code at Nodes to store the created nodes as children of a new or existing node- OK. TO SEARCH TEXT Query- New- Text Search Resources for Topic 7 Within NVIVO 8: Help topics coding, autocoding Teach Yourself NVivo 8: Tutorial 5 Nodes and Coding, Richards, 2005: Chapter 5 on coding Bazeley, 2007: ch 4 on coding and pp on searching. Her advice on formatting for autocoding is on her website under resources- technical resources. Go to RMIT University Page 12 of 25

13 Topic 8: Creating cases and attributes A case here is a unit of analysis (e.g. a person, or an organisation). If you plan on doing any comparison in your study, create case nodes. You can bring in sources as cases, or can create them as cases later. NVivo stores information about cases as attributes. Only case nodes can have attributes but apart from that fact case nodes are like nodes in most respects. They can have memos attached, be put into sets etc. IMPORTING SOURCES AS CASES CREATING CASE NODES INSIDE PROJECT ADDING SOURCES TO AN EXISITNG CASE (e.g. two interviews per participant/ cases are organisations and you have lots of sources for each organisation) When importing an internal select Code sources at new cases... Nodes- cases- new Keep the sources separate (so you can compare pre and post interviews, or data from CEOs with data from line workers), and code the new sources to the relevant case node. Creating sources as cases automatically creates a casebook. You can create case attributes (e.g. gender) with their values (e.g. male and female) and record them in the casebook. Values of attributes can be numbers, strings of characters, Boolean (yes/no) or dates. If you know what attributes you need, consider creating them early and consistently storing the information as cases are introduced and coded. They can be changed later. CREATING ATTRIBUTES CREATING VALUES RECORDING ATTRIBUTE VALUES FOR A CASE Classifications Attributes - New (In attribute property box) Values add - OK Open Casebook click required cell - use drop down arrow to insert existing values or click on it to make value cell editable and type in new value If you have demographic data stored as a table (e.g. in Excel) you can import a Casebook. For detail on this, see Bazeley 2007 pp Be careful that the names of cases in your table are the names used for your sources. Resources for Topic 8 Within NVIVO 8: Help (under search) about cases and attributes follow to Importing Cases and Attributes for information about formatting and importing data from SPSS or Excel. Teach Yourself NVivo 8: Tutorial 3 Managing data: Cases, Attributes and Sets Richards, 2005: Chapter 3 for advice designing the use of attribute data, and on the ways of grouping and focusing on data. Bazeley, 2007: ch 6 RMIT University Page 13 of 25

14 Topic 9: Using find A common use for find is when you want to locate project items and can t remember where you put them, for example, a particular node that may be nested three deep in any of a whole forest of trees. (Ideally we don t have such unwieldy projects, but in real life of course we do) FINDING PROJECT ITEMS Enter name (or partial name and * or?) in Look for on the Find bar. (The find bar is at the top of list view) GATHERING PROJECT ITEMS OF A PARTICULAR TYPE SAVING FINDS Use intermediate or advanced tab on the Find bar Select items you wish to save, RMB, createcreate as set Resources for Topic 9 Within NVIVO 8: Help topics Find Teach Yourself NVivo 8: Tutorial 8 Asking Questions Richards, 2005: For advice on using software searches in qualitative research, and warnings about misusing them, go to Chapter 8 of Handling Qualitative Data. Bazeley, 2007: pp 134 on find RMIT University Page 14 of 25

15 Before we move on: Thinking about queries NVivo provides different types of queries that enable you to question your data find patterns and pursue ideas. You can save queries, re-run them through new data and track the evolution of results. You can create the following types of queries: text searches, word frequencies, coding queries, matrix coding queries, compound queries and coding comparisons (the last is not covered in this workbook.) Some projects use no queries, some a few simple queries, others many and complex. Here s the NVivo help s general and wise advice for queries: 1 First, state the common-language question you wish to ask. 2 Then select the type of query you need. The question who was it who kept talking about excitement? requires a simple text search. Did she really use the word that much? needs a word frequency query. Do people who talk about excitement also talk about volunteering abroad? is a coding query. Do men and women have different images of volunteering asks you to find the stuff you have coded in the images of volunteering tree and more or less cross-tabulate it by the attribute gender, so it is a matrix coding query. Does the word community come up in or near stuff I coded as civic responsibility requires you to look for text and check out its proximity to coded material so it is a compound query. 3 Restate your question in terms of the processes it requires. Step logically through each process. Quite often, you need to tell NVivo that the search is limited to only some sources, or only some values of an attribute. (For example, you probably don t want your matrix table to include cases whose gender is coded as not applicable or unassigned ). 4 Run your query and then look at the results. Carefully state the answer to the common language question that was your starting point, accurately recording what it tells you. And then (Pat Bazeley s advice here) WRITE DOWN WHAT YOU HAVE LEARNED so you don t find yourself staring at the results tomorrow wondering what on earth they mean. 5You can run a query without saving, or save the information required to set up the query (by clicking the Add to Project checkbox). When you run a query, the results are displayed in the detail view. You can set up the query to save the results, or you can preview the results and later save them. The default location for saved results is the Results folder. You can move results into a free node, tree node or case folder. You cannot code at nodes in the Results folder, but once you move the results into a tree or free node you can code on from it. RMIT University Page 15 of 25

16 Topic 10: Text Searches Text searching to find material is useful for locating the sources containing a particular word, phrase or image. You can select the spread of context in which to display the found item (by itself, in a narrow or broad context) or you can spread the context after the search is done. If you wish, you can set up the query to automatically code the results into nodes, or you can code them later. As a way to speed up coding text searching is very useful in very limited circumstances: e.g you can readily collect all mentions of Ventolin using a search but you won t get all mentions of my asthma medicine that way, A text search might be quite useful to get a sense of whether it s worth doing more detailed exploration. (If only three people mention Ventolin, but all say things that warrant further thought, it s probably worth going back through all the interviews looking for comments on asthma medication ) NVivo s default settings for context for text: Narrow ads 5 words either side, Broad ads the surrounding paragraph, Entire Source adds all content. TO SEARCH TEXT TO ALTER THE CONTEXT OF FINDS TO STORE RESULTS AS YOU CREATE THE QUERY TO STORE RESULTS ONCE PREVIEWED Query- New- Text -Search In Query options spread coding In Query options Results -options Select results, RMB store query results Resources for Topic 10 Within NVIVO 8: Help topics Text search Teach Yourself NVivo 8: Tutorial 8 Asking Questions Richards, 2005: chapter 8 on searching Bazeley, 2007: pp on searching RMIT University Page 16 of 25

17 Topic 11: Word Frequency queries You can check the most frequent words in particular places (e.g. certain sources or certain coding) WORD FREQUENCY QUERY New query-word frequency- select context and display options- run Resources for Topic 11 Within NVIVO 8: Help topics Word Frequency Queries RMIT University Page 17 of 25

18 Topic 12: Coding queries and compound queries Is my brilliant insight that everyone of this kind talks about this topic supported by the data, or have I been seduced by one memorable participant? To check this I might want to see what female participants in a particular age group say about a topic. If the only person I find is Anna, then my insight is clearly the result of seduction. If all or most of the participants of this type turn up when I run the query, my insight is worth further exploration. TO CREATE A CODING QUERY Query - New Coding query -choose simple or advanced. If advanced, define criteria and set operators You can build complex queries that combine the results of searching with coding queries- the trick is to think carefully about what you wan to know. Maybe, for example, you want to know if participants of a particular kind who talk about one topic also talk at the same time about a second topic. Task: Run a compound query TO CREATE A COMPOUND QUERY Query - New Compound Query. Select subquery type for the first part of the query, select options for proximity, select options for the second subquery type run Resources for Topic 12 Within NVIVO 8: Help topics Queries- coding queries, compound queries Teach Yourself NVivo 8: Tutorial 8 Asking Questions Richards, 2005: for advice on using software searches in qualitative research, and warnings about misusing them, go to Chapter 8 of Handling Qualitative Data. Bazeley, 2007: pp on coding queries p171 compound queries RMIT University Page 18 of 25

19 Topic 13: Running a matrix query and showing matrix results visually Looking for regularities in one s data is a standard technique. Even though most of the time we are very cautious about extrapolating any pattern in our data to a wider situation, we want to know if there are patterns. The matrix query is a powerful tool for the kind of sorting and sifting that once was done by putting pages in piles (and risking having the baby eat a quotation!) The Matrix query tool enables you to do comparisons. It s called a matrix coding query because matrices are made of nodes that code data. The items you specify for rows or columns are nodes, and the cells of the table will contain data coded at the combination of those nodes you specify. You specify what nodes or attribute values make the columns and the rows and the relation between them that builds the matrix. As for any query, you can also specify the scope of data searched. For most qualitative research, the critical purpose of building such a table is to return to the data and try to understand and interpret the pattern. For any matrix, you can view all the content that belongs in each cell and work with the data, because the cell is a node. You can specify what you do with the result - save it as coding so it can be used in another search, and/or add it to the project so it can be re-run. You can also display patterns with charts in NVivo8. Again, you can move from the display of the pattern back into the data.. TO CREATE A MATRIX QUERY CREATE A MATRIX chart Query, New Matrix coding Query, (useful here to add to project ), define your rows and columns (don t forget to add to list ) select operator (e.g. and ) set your query options, run Tools - charts Remember what you are counting- is it cases or references? And remember that the point of the matrix is the quality of difference. But don t completely neglect the value of seeing the numbers, and doing the basic arithmetic to enable you to describe what s going on in your study. (Think about the statement few women but many men said. In a sample of 23 men does many' start at 11½ in your view, or at 15, or somewhere else? It may be useful to be able to say 15 of the 23men compared to 3 of the 11 women ) Resources for Topic 13 Within NVIVO 8: go to Help, Index, Matrices, charts Teach Yourself NVivo 8: Tutorial 9 Exploring Patterns in Matrices Richards, 2005: Chapter 9 Bazeley, 2007: Chapter 6 RMIT University Page 19 of 25

20 Topic 14: Using Relationships Coding at free and tree nodes has indexed content as being about a category (a topic or an analytic concept perhaps). Relationship nodes store material, but instead of being a category it is a statement (and possibly evidence) of a relationship between two items in your project. The material under the category Bill can be linked to material under the category Jane with the statement is married to. This relationship might not have anything in the node. But a relationship "encouragement" is important to "motivation"' might have in the node all the material providing evidence of a link between these two concepts. Relationships may help with patterns where you are thinking about causation in a nonquantitative sense. CREATE A RELATIONSHIP TYPE ADD A RELATIONSHIP CODE FOR A RELATIONSHIP (navigation view) Classifications, New, (list view) Relationship type, New, select direction (In navigation) Nodes, relationships, New relationship. Select items for the to and from fields, select type. In a source, select the material that is evidence of the relationship and code it at the relationship node Resources for Topic 14 Within NVIVO 8: Help topics, Relationships Teach Yourself NVivo 8: Tutorial 6 Relationships and Other Nodes Richards, 2005 ch. 5 Bazeley, 2007 pp116-8 RMIT University Page 20 of 25

21 Topic 15: Running reports As analysis develops, we want ways to show what we are finding. There are many ways in which NVivo can help you report what you are thinking and demonstrate the rigour of your approach. You can copy and paste material, or export entire items. You can generate reports for exporting. You can export to a Word (or.txt or.rtf) file. It s easy to copy and paste from the NVivo project to an outside document or to export items wholesale. If you are thinking of exporting the contents of a node so you can write about them, you can simply copy and paste from the detail view using icons or fast keys. But note that while hyperlinks remain live in pasted text, if you copy and paste from an NVivo source to a document in Word, annotations are lost. So if you want the full detail of the text, it is better to export the item. EXPORT A NODE Select the node you wish to export, Project, Export Item, select the options you want, OK, Save As dialogue, specify name, location and type of file, OK. (In the saved files Annotations or See Also Links will appear as endnotes) You might want to convince readers that you have covered your data thoroughly. You can export the list view of sources, or create a report of sources or nodes. The second way gives a formal report that appears on the screen as a formatted document, and can be printed. If you want to save and edit it, it can be exported as a Word file. But be a little careful here- large reports take time. Don t request a coding report of all sources and all nodes! You may not want such a formal report. NVivo gives you in the list view which includes the extent of coding of the items in the list, and you can export this (it comes up as an Excel file). CREATING A REPORT EXPORTING A REPORT EXPORTING A LIST OF ITEMS Tools, Reports, select the type you require, define format options, OK. In the report viewer hit the export icon (Navigation view) click on the relevant areas to export, select the items to export, Project, Export list or RMB, Export list Resources for Topic 15 Within NVIVO 8: go to the two topics in online Help (Using the software and Working with your Data on Reports) Online Tutorials: go to Teach Yourself NVivo 8: Tutorial 10 Reporting and showing your project. Richards, 2005: Chapter 10 has advice on what you might use the reports for- particularly on ways of ensuring that your report is convincing enough. RMIT University Page 21 of 25

22 Topic 16: Models Models are another way to report on data and convey developing ideas. They can also help generate the ideas. Models can be created and items added to them at any time. You can place in them o shapes you can name, color and resize o project items, including relationships o connectors between items. Connectors can show direction or simply association. As you move towards a final report, models can show the project in part or as a whole MAKING A NEW MODEL Model, New You can ask to see a project item s associated items and have them brought into the model. Models are dynamic (If you bring in a node, you can get back to the content by clicking on the item in the model). This is helpful for thinking, but takes up space. You can save any model as a static model no longer live to the data. This helps you show the progress of your project. The model made last month may contain nodes you have since merged or deleted, or relationships that proved insignificant. They will not appear in a dynamic model, since they are not in the project. But they are still there in a static model, and you might want to save these models so you can show readers how your thought evolved. You may want to write a memo for each static model saved, summarizing why you saved it, and what it shows. SAVING A MODEL AS A STATIC ITEM TAKING A MODEL OUT OF NVivo Detail View of the model you wish to save as static Project menu Create, Create As Static Model, name and describe model, Select items you want, Edit, Copy. Open location in destination file a word document/ PowerPoint slide etc, Paste. Resources for Topic 16 Within NVIVO 8, go to the two topics in online Help (Using the software and Working with your Data) on Models Teach Yourself NVivo 8: Tutorial 4 Editing and Linking includes comments on the early use of models. Tutorial 10 Reporting and Showing Your Project has more detail. Richards, 2005: Chapter 7 Bazeley, 2007 pp34-6 for the basics, 95-7 for material on the advantages of modelling a case, and173-5 for playing with models. RMIT University Page 22 of 25

23 Finally: Some Resources NVivo is a tool that you use to suit your own project and your own analysis. It enables you to keep your data and your ideas together in one project, in ways that should help you move from sources to conclusions. You can use it as a filing system alone, but contemplate going further than this. The resources below will help you think about other ways to use NVivo. On using NVivo for analysis In addition to the resources already listed o The CAQDAS program has a useful website on using programs like NVivo. Check out their REQUALLO pages. o Marshall, H ' do we do when we code data?' Qualitative Research Journal 2: pp56-70 is my own take on some of the practical issues involved in coding. It s available electronically via INFORMIT or through the website of The Association for Qualitative Research o Sage publications runs Methodspace, which offers access to articles plus blogs, facebook, twitter etc. Making use of software in research design Di Gregorio, S and Davidson, J (2009) Qualitative Research Design for Software Users Open University Press, Maidenhead, Berkshire On using NVivo to demonstrate rigour o Bringer, J., Johnston, L.H., and Brackenridge, C.H (2006) 'Using Computer Assisted Qualitative Data Analysis Software (CAQDAS) to Develop a Grounded Theory Project', Field Methods, 18(3): o Bringer, J.D., Brackenridge, C.H. and Johnston, L.H. (2004) 'Maximising transparency in a doctoral thesis: The complexities of writing about the use of QSR*NVIVO within a Grounded Theory study', Qualitative Research 4 (2): For more on how students have used NVivo, see o Siccama, C J; and Penna, S (2008) Enhancing Validity of a Qualitative Dissertation Research Study by Using NVIVO Qualitative Research Journal 8:2 o Consultation Robertson, S (20080 The Qualitative Research Process as a Journey: Mapping Your Course with Qualitative Research Software Qualitative Research Journal 8:2 Those who have attended a workshop with me are entitled to one free ed/phoned conversation about their project. Good luck with your research! Helen.marshall@rmit.edu,.au RMIT University Page 23 of 25

24 Phone: RMIT University Page 24 of 25

25 NVivo 8 System Requirements Component Minimum Recommended Processor 1.2 GHz Pentium III-compatible processor 1.6GHz Pentium 4-compatible processor or faster Memory 512MB RAM 1GB RAM or more Display 1024 x 768 screen resolution 1280 x 1024 or higher Operating System Hard Disk Other Microsoft Windows 2000 Professional Service Pack 4 Approximately 1GB of available hard-disk space Windows XP Service Pack 2 or later, including Microsoft Windows Vista Service Pack 1 Approximately 2GB of available hard-disk space Internet connection NVivo 8 is designed to operate on Microsoft Windows natively. If you are running the software on a virtual platform on a Mac, different system requirements may apply. RMIT University Page 25 of 25