ATLASti_Newsletter_2011
Contents
1. INSIDE ATLAS ti The QDA Newsletter e o l i l i i s In This Issue This issue of INSIDE ATLAS ti has a preview of the upcoming version 7 0 release For a list of some of the most important new features changes and powerful ad ditional functionality you can expect from this latest major release read more here In Research with ATLAS ti Corey Abramson of the UC Berkeley Center for Urban Ethnography writes about Qualitative Research in the Positivist Behavioral Tradi tion Read the article here New Publications is a new section that informs our readers of select scholarly publications that have come to our attention Read more here We highly recommend an important new book that is scheduled for release in December Dr Susanne Friese Qualitative Data Analysis with ATLAS ti Read more here And last but by far not least we announce the pub lication of the new updated and expanded edition of the ATLAS ti User Manual Read more here Newsletter 2011 3 December 2011 We hope you enjoy reading this new edition of INSIDE ATLAS ti and invite you to send us your feedback newsletter_feedback atlasti com Happy coding Jorg Hecker Director Business Operations Xmas Sale All licenses 20 Off To give back something for the fantastic development of the past year this holiday season we offer ANY NEW LI CENSE including student licenses for a full 20 off dur ing the month o2. December 2011 That context requires that contend with classical positiv ist behavioral concerns about bias error representation inferential logic and reliability Many of the individuals and teams work with use ATLAS ti in similar organi zational and disciplinary contexts Some agree with the underlying logic of positivist behavior research others find it repugnant There are myriad critiques and respons es some of which list at the end of this article Still regardless of where they fall in these debates researchers acknowledge that there is a comparative dearth of litera ture relating CAQDAS to these concerns This article is not a normative assessment of various epistemologies or the conditions under which one might adhere or deviate from them My intent is not to provide prescriptive statements about what constitutes proper research Rather ad dress in a limited way how CAQDAS relates to concerns with the possibility of error in inferred associations under the positivist behavioral tradition and show how some existing tools in ATLAS ti can be used by those operating in this tradition CAQDAS and the Problem of Inference Two of the most common questions get asked during workshops and consulting are 1 How do know if an association between codes is real and 2 How do deal with relationships that are real but do not show up in my examination of code associations Both questions are concerne
3. View In ATLAS ti 7 you can work on multiple documents at the same time and side by side Imagine the possibilities Com pare different versions of the same document directly draw connections between different segments or synchronize multimedia and text documents with one another Newsletter 2011 3 December 2011 atlas ti Document regions can be freely arranged via drag amp drop import e g comments tags and the added security of The operation is smoothly animated full optional project encryption Networks Reloaded and Virtual Links The visual aspect of working with network views one of ATLAS ti s particular strengths from the start has been significantly improved further A new node display with realistic drop shadows and full node dragging individually resizable nodes and node hiding make visual based work ing more powerful Add to that a new animated layout flicker free resizing real zoom for greater freedom as well as the use of virtual links e g PD code links and view persistence on node and network level for a great new work experience and added flexibility j l f Multi Document View Three primary documents dis played side by side associated with Up to four documents of any type can be displayed and worked with side by side including their margin areas This enforces an in depth understanding of the data It also makes transcribing media and analyzing associated l documents
4. as well as live at numerous events all over the world Facebook http facebook atlasti com YouTube http youtube atlasti com Newsroom Visit the ATLAS ti Newsroom for a comprehensive overview of all media channels and press information Newsroom http newsroom atlasti com e ee e eee ees Newsletter 2011 3 December 2011
5. child s play A transcript can be activated from its associated media file clicking the red dot Cross doc New Network View engine with real zoom and excel ument drag amp drop of brackets lets you create hyperlinks lent printing and navigate them without losing context Fast Flexible and Secure Document Virtual Links e g display of codes connected to their Management Managed Documents primary documents is now possible without a need to populate the workspace with the quotations they are actually linked to ATLAS ti 7 s powerful new managed documents con cept ensures referential and content integrity to make users life easier You now have a choice if you want to keep your project files stored in their original location as before or if you want ATLAS ti to handle all document management tasks for you safely and securely The man aged document concept also includes a public library that contains shared material for collaborative work Finally excellent full printing capability and high resolu tion image export of network views give you countless additional options for reporting and added usefulness all around rtf with Tables You can retain the project document structure you re used to or convert your existing HUs to managed docs ATLAS ti 7 imports doc docx and rtf fully formatted This sophisticated document management also gives including tables Whatever your source material looks like you i
6. explained by the coming of the summer months which increases both ice cream consumption and swimming If we agree to this sort of logic the association between ice cream and swim ming deaths is a statistical artifact It is what can be a called a type error or false positive There is a formal mathematical association between two concepts or codes but the association is ultimately spurious or non substan tive and explained by something else another measured or unmeasured variable concept or code in this case time of year It is also possible that two concepts or the codes that represent them are related in a crucial way but that codes do not come up as formal associations in queries This is referred to as a type II error of false negative In this instance there is a real relationship between two things in the world e g time of year swimming and likelihood of drowning but this relationship is not easily reflected in the data Using the example above one might search for time of year summer and drowning but the resulting query does not yield useful results Or perhaps there is no existing code or concept to refer to what is assumed to be the explanatory factor time of year Here the lack of formal association in the software does not map onto the substantive association between time of year and drowning that exists in the larger universe of investiga tion In addition to the type one error the false positive the
7. a Analy sis with ATLAS ti by Dr Susanne Friese is the very first book designed to guide you step by step through your research project using ATLAS ti In the book you will find clear practical advice on preparing your data setting up a new project in ATLAS ti developing a coding system asking questions finding answers and preparing your results ATLAS ti Manual More information and pre order is available from Sage UK http www uk sagepub com books Book235168 tabview title Sage US http www sagepub com books Book235168 Read some early reviewers voices and download a flyer for distribution here http www atlasti com frieseqda html The full ATLAS ti 6 Manual has been significantly updated revised and expanded Download it from http manual atlasti com It contains full documentation of the program and useful information for special application scenarios Those just getting started may prefer the light version of the manual also referred to as the mini manual It makes the perfect companion for the Quick Tour and its pre coded sample project that come bundled with ATLAS ti The mini manual has answers to many fre quently asked questions and provides useful tips on opti mal project setup and management basic information on coding data analysis and visualization with the network view function It is available for free download from http www atlasti com uploads media m
8. accounts are typically constructed in a non investigative context More specifically the aim of the project was to show whether there are any dif ferences in the narrative accounts provided by different offenders classified according to the type of crime they committed and their experience in the area of deviance All interviews were analyzed using ATLAS ti and taking the Evaluation model of W Labov as a reference Partici pants were divided into three groups according to their criminal experience ie professionals intermediates and amateurs and into four groups according to the crime they committed ie homicide robbery and theft and crimes related to drugs and fencing The results show that there are important differences in the way a crime is reconstructed by offenders with different criminal history The paper also proposes solutions and innovative strate gies to deal with quantification and qualitative data Newsletter 2011 3 December 2011 atlas ti Patrizi Patrizia 2005 Deviant Action and Self Nar ration A Qualitative Survey through ATLAS ti Journal for the Theory of Social Behaviour Volume 35 Issue 2 pages 171 188 June 2005 This paper has its roots in qualitative analysis of accounts produced by an Italian serial killer Theoretical references are related to symbolic interactionism and its develop ments in the field of psychology and criminology The whole contribution is aimed to twofold purposes A to p
9. apparent lack of a relationship when one exists in the universe outside is a type two error a false negative I am often asked questions that knowingly or unknow ingly touch on this issue of inference People ask how do make sure my data and coding capture what is going on how do make sure a relationship isn t just made up and what if there is something know is going on in the world but isn t showing up in my data set Often the underlying assumption is that there is a simple answer and that their difficulty is the result of a lack of technical proficiency with ATLAS ti The truth is that while technical proficiency is necessary to address these problems par ticularly for large qualitative data sets these are complex methodological questions They speak to the palpable tension of how to balance type and type II errors in the construction and analysis of data There is no single com mand i e right click minimize type errors or technique atlas ti that will dissolve this problem There is no one size fits all solution Even within the methodological logic outlined above which is only one of many responses to epistemo logical issues such as how to manage truth and corre spondence the researcher and analyst must make hard decisions about how to balance and manage type and type Il errors in specific projects While there is no one size fits all solution in grappling with these issues in b
10. ative research to account for and incorporate observations the analyst didn t expect The ultimate concern isn t simply how good the initial coding scheme is but whether or not the emerging analysis which relates to both existing theories and the real world is correct in the end Tools for Reducing Type II Errors in ATLAS ti Given this general strategy for dealing with error there are a number of basic and advanced tools in ATLAS ti that Newsletter 2011 3 December 2011 atlas ti use to this end Each of these are covered well by exist ing articles and documentation in ATLAS ti The idea is to think about these not simply as strategies for exploring data but for decreasing type II errors Each of these can be used to help patterns you may have missed Rereading data Control f allows you to search for strings of text in a way that may not line up with your intial coding The object crawler does this as well and provides more options Inthe query tool you can use the follows and precedes operators instead of limiting yourself to Boolean and or not Xor searches This can help you in identifying chronological patterns Using quantitative outputs e g the code primary document table could broaden the scope in looking for a relationship between codes You could use this to see for instance if and how many times two codes are in same primary document Co occurency table explorer This
11. ctives to be an important strength of current generation Computer Assisted Qualitative Data Analysis Software CAQDAS The software accommodates phe nomenologists interpretivists grounded theorists positiv ists and a whole host of other epistemic orientations ATLAS ti does not require the analyst or team to impose an exclusively top down deductive logic or an emergent inductive logic but allows the deployment of both logics either alone or in combination Because of this flexibility CAQDAS workshops are one of the few venues where it is not odd to have political scientists sociologists public health scholars rhetoriticians anthropologists and market researchers in the same room Often the key common denominator is that they recognize the software as useful for their research or inquiry even if they agree on little else Consequently when teach ATLAS ti do not pros elytize or try to advocate for a particular methodology or epistemic perspective but rather show how the software can be used as a flexible tool to accomplish individuals particular goals always believe that the best practices for using software like any tool for social research are those practices that work best given the user s philosophy and aims Nonetheless as a sociologist that does work on inequality health and policy at a large research university operate within a particular scholarly and organizational context Newsletter 2011 3
12. d alone and full screen Pens la emen Quien Com Memes Needs Anshan Toen ext Wate elions Hep PU 4 A S RRE S o w Ce New Multimedia engine View video PDs frame by frame for in depth scene coding and multi dimensional analysis All codes display in the margin area just as nature intended Multi dimensional analysis Create image PDs from dis played frame for in depth scene coding Virtual Reality Margin Metaphor Throughout Most intutive user interface Software users not only ATLAS ti still prefer to simulate the non digital world even in this digital age Smartphones offer virtual bookshelves from where to pick your eBooks Pages of eReaders can be turned over as if they were real pages Music apps let you input rhythm patterns via drum pads and melodies via virtual keyboards or virtual In nearly every field there exists a time honored and well proven standard way of interaction with your material The most natural environment for the qualitative re searcher is the margin area Audio and Video now share ATLAS ti s basic metaphor for presenting and working with documents the margin area Text PDF images audio video all utilize this powerful workspace Printing your source materials with all its margin informa tion not possible in any other QDA product is naturally standard procedure for ATLAS ti In ATLAS ti 7 this option will also be available for PDF and image files Multi Document
13. d with the representation of a social or linguistic process in a data set and how it maps on or fails to map atlas ti on to a reality outside It is a component of a set of clas sical problems in analytical inference how does one get correspondence between recorded data and the phenom ena they purport to record How can theory or model be made line up with the phenomena they explain And on a related note how does one minimize the possibility of error or miss correspondence In essence the how do we know questions are complex philosophical problems underwritten by particular logics of inference In the posi tive behavior tradition the overarching concern is how can we deal with the lurking specter of type errors false positives and type II errors false negatives in research generally and qualitative research more specifically Dealing with problems and theories of inference and cor respondence is a huge field with volumes of books and articles addressing possible responses My contribution in this article is much more meager do three things 1 address how one might examine the relationship between two or more codes 2 examine how this process relates to the concern of type and type II errors and 3 discuss some of the new tools in ATLAS ti that can be deployed to help minimize inferential errors Defining Codes and Coding Before examine how to examine the relationship be tween codes i
14. ent license Newsletter 2011 3 December 2011 atlas ti Qualitative Research in the Positivist Behavioral Tradition Resources for Addressing Type I and Type II Errors in Code Associations Using ATLAS ti Corey M Abramson Corey M Abramson http cmabramson com is a PhD candidate in the Department of Sociology at the University of California Berkeley and a Research Analyst at the University of California San Francisco s Institute for Health Policy Research His research uses a combination of quantitative qualitative and formal theoretical methods to examine the mechanisms that link social inequality to the health of populations individuals and the human body more generally Corey currently manages UC Berkeley s interdisciplinary Center for Urban Ethnography CUE where he collaborates on projects with scholars from various fields works on methodological issues and directs an undergraduate training pro gram in qualitative methods He is also the lead instructor in Computer Assisted Qualitative Data Analysis CAQDAS for the UCB Social Research workshops where he provides methodological training and consulting to faculty graduate students and research teams from across the country Introduction When teach workshops on data analysis in ATLAS ti get to interact with people from a wide spectrum of disci plines and philosophical orientations consider this ability to accommodate a host of methodological and epistemic perspe
15. etic Teamwork Enabled Out of the Box ATLAS ti gives you the freedom to work alone or in a team without the need for special licenses add on mod ules or servers Collaborative work is possible with any in stallation and license type The new Managed Document feature supports central storage for team projects and helps you keep your project data organized in a central public library locally or remotely Also each data source is now identifiable via a unique ID This means the position of the data source in the list of documents is no longer an issue when merging multiple projects Each team member can now freely adjust the sequence of documents to fit individual analysis needs and preferences Administrators will enjoy new support for silent unattended installations New Analysis Options A number of new analysis options are still in the pipe line A description will be available in the next edition of INSIDE ATLAS ti btain an upgrade for your existing license simply ittp www atlasti com order html select ade and enter your license key when asked vely simply contact our support center via support atlasti com and send your license key des are available for all older versions Up prices are generally approx 50 of the price ew license Upgrades can only be obtained for ne type of license Student licenses cannot ded but if you still have full time student you can simply apply for a new stud
16. f December Simply enter the following coupon code on the shopping cart page 2F9 NQS SJR Important to note Not only will you receive the full pack age of ATLAS ti at this lower price you will automatically also get a FREE UPGRADE to v 7 0 once it is released in early 2012 Thank you enjoy and happy holidays from ATLAS ti atlas ti ATLAS ti v 7 Preview ATLAS ti 7 is only a few months away its release scheduled for early March 2012 Time to take a more in depth look at what is coming in this major new release There are many reasons to look forward to this next milestone in profes sional qualitative data analysis A detailed preview of select new features can be found here http downloads atlasti com v7preview v7preview_201111 pdf The guiding princples of the innovations in ATLAS ti 7 can be described as centered around increased usability stream lined productivity and revolutionary new functionality Powerful Video and Audio Analysis ATLAS ti 7 features state of the art analysis of multimedia files This makes ATLAS ti the premier tool choice for anybody working closely with video and audio material such as media researchers educators video analysts musicolo gists and scholars in a host of areas where media analysis is required Waveform cues and frame previews overcome the navi gational blindness of the old approach and give you highly precise navigation Videos can be viewed embed ded stan
17. ger meta categories which facilitate the production of a model theory or explanation Some methodologies e g grounded theory use this form of coding exclusively others verbally discard it as subjective In practice the inductive generation of codes is a common aspect of most qualitative research even in the most rigidly positive behavioral disciplines since to ignore unforeseen patterns would be to miss out on one of the key strengths of qualitative research The second mode of generating codes is deductive This sort of coding is common for positivist behavioral re search clinical methodologies the extended case method and numerous other approaches found in social science and policy disciplines Here researchers begin a project with an understanding of specific existing theories that purport to explain why things happen the way they do in the world They want to account for these in their analysis of qualitative data After doing a project grant proposal literature review etc these researchers will typically go through and generate codes based on existing possible and counter factual explanations for the social phenom enon they are studying refer to these codes which are typically generated before fieldwork or data analysis as Newsletter 2011 3 December 2011 atlas ti deductive codes Deductive codes operationalize existing theories explanations or categories from prior empirical and theoretical inquiry As wit
18. h inductive codes these can include everything from demographic categories such as age to abstract theoretical constructs The codes may not correspond to or be applied to the data but they exist initially to orient a researcher to an analytical frame If they correspond to the data even in part that is a key finding If they do not correspond at all e g they are use less in explaining a social phenomenon that is a finding as well Generating at least some codes deductively also saves a great deal of time in the process of recoding The most common practice in sociology and related dis ciplines is currently to use a combination of inductive and deductive codes Deductive codes are generated based on prior understandings of a topic existing theories and hypothesized explanations that purport to explain a specific research puzzle Inductive codes are generated to correspond to the unforeseen patterns and occurrences that present themselves during the course of produc ing and analyzing qualitative data Since unlike survey methodologies the range of responses or observations in qualitative research are not limited to a drop down menu pre generated at the outset of a project inductive codes are necessary to account for observed occurrences or patterns that were not in the initial analytic frame De ductive codes are generated using a top down approach e g starting with explanations and seeing if they map on to data Induct
19. iniManual_v6_ 2011 pdf Highly recommended Newsletter 2011 3 December 2011 atlas ti Talk To Us We Love To Hear from You We hope you enjoyed reading this issue of e A INSIDE ATLAS ti Your Quarterly Newsletter And we welcome your feedback and suggestions Newsletter Feedback Help to ensure that we fill future issues with information and news that are relevant to you Simply use the feedback link to tell us what you liked and what you would like to read about in future issues newsletter_feedback atlasti com Request Suggest Program Features Your suggestions and requests are an invaluable source when it comes to further improving our software Help us identify your needs by telling us about product features that would make your research efforts easier or more efficient http suggestions atlasti com Newsletter Settings To update your preferences and contact data to subscribe or to unsubscribe from our newsletter use the link below Also you can comfortably recommend INSIDE ATLAS ti to friends and colleagues from this page http newsletter atlasti com lists p preferences Social Media You don t have to wait for the next issue of INSIDE ATLAS ti for the latest news Just connect with us on Facebook or visit our YouTube channel and you can be in touch any time We look forward to all your comments And rest assured We continue to be avail able as always through our forum mailing list website
20. is a new and extremely powerful tool for seeing patterns in your data See the new newsletter article by Ricardo B Contreras here http downloads atlasti com library contreras_ nl201108 pdf You can use network views to examine the associa tion of codes from different vantage points Newsletter 2011 3 December 2011 Conclusion The questions how do we know if an association is real and how do we deal with real associations that do not show up in a query are not technical questions They are fundamental issues tied up with epistemological concern like how do we know what we know As such there is no simple answer and the most useful or palatable solu tion varies between and within disciplines While criticisms of the positivist behavioral tradition abound particularly in qualitative research this is context or tradition in which many of us operate Without offering prescriptive statements about the relative utility of different episte mologies have discussed a problem fundamental to research in the positivist behavioral tradition dealing with and minimizing errors in association showed how this relates to CAQDAS and the specific tools that ATLAS ti has for this purpose It is my hope that even if one wants to contest this logic having a better understanding of it and the way ATLAS ti s tools might be implemented can strengthen their own inquires atlas ti Publications In this section we let yo
21. ive codes are based on a bottom up approach beginning with the data and building up The extent to which one uses inductive versus deductive codes or even uses codes is a function of their methodological tradition and epistemic choices Codes Errors and Inference There are two types of relationships between codes key concepts themes or variables in a qualitative data set First there is a mathematical formal relationship between the codes in ATLAS ti This formal relationship determines what comes up in the window when one does a query or produces advanced output like co occurency tables These codes can be related by boolean links semantic links or proximity links The ATLAS ti manual s discussion of the query tool explains these links in detail In short these links allow the researcher to see if codes come up together come up alone or are following preceding overlapping one another There are more complex asso ciations that are possible but that is the basic idea Newsletter 2011 3 December 2011 The second type of relationship is a substantive empirical relationship Some of the formal associations that show up in ATLAS output may indicate a spurious relationship The classic example in elementary statistics is the positive correlation between ice cream consumption and drown ing The common argument here is that the association is spurious explained by an unmeasured explanatory variable time of year Both are
22. lished in 2005 this article aims to identify tourism trends in the Asia Pacific region The qualitative analysis sketches a holistic conceptual network of relationships among trends whereas the quantitative analysis visually presents the trends commonalities or differences among four subregions Northeast Asia Southeast Asia South Asia and Oceania This article differs from previous studies in that it counts the frequencies of quotations associated with codes rather than the commonly used keywords Quotations are composed of sentences and paragraphs that provide a deeper and richer understand ing of the context And as the title implies it visualizes the relationships among trends and between trends and sub regions Through visualizing the qualitative data holistically and individually this approach is intended to see the forest for the trees De Gregorio Eugenio 2009 Narrating a crime Con texts and accounts on deviant actions International Journal of Multiple Research Approaches Vol 3 Issue 2 177 190 This article summarizes the main findings of a study on the narrative construction of deviant actions which fo cused both on the contents and the structures of the nar rative accounts provided by criminals The qualitative re search was conducted by collecting narrative interviews These involved 34 individuals detained in the penal in stitutions of Regina Coeli and Rebibbia Rome and were aimed at revealing how
23. ncreased performance through one time doc to rtf ATLAS ti can handle it We have now decided to move conversion revision management optional metadata to a new rtf model but do so without jeopardizing older Newsletter 2011 3 December 2011 Page atlas ti project data That means you can enjoy the full width of formatting options including tables for all rtf documents that are added to your projects in v 7 Word Clouds Wordcruncher A staple of many Web 2 0 applications and already avail able as an output format via the XML stylesheets since v 5 ATLAS ti 7 now integrates Clouds even more seamlessly into the program core Clouds for primary documents and codes are now part of the Wordcruncher interface and make many analytical steps more intutive GUI Changes The ATLAS ti user interface has undergone numerous subtle but extremely usability enhancing changes They will increase your productivity without interupting your workflow For example Flexible side panels unify central functions and give speedy access to frequently used functions and tools Q cuss tg Cotter oo se anasi Wort Freer Cs oaa ee a z name of suspect 1 j i umdkia I el 1 by whon are on_suspected oY investigati Jack Word Clouds Open coding dialog auto completion makes coding even faster and easier And last but not least subtle anima tions make use of many common tools more pleasant and aesth
24. o minimize type II errors by generating codes deductively at first creating codes inductively to correspond to new findings coding densely with all relevant codes then go ing back using the query tool to rule out false positives look at formal associations as something potentially true then decide if this is a false positive not by computing the probability of getting such a result due to chance but re examining the event or phrase in ATLAS ti typically the quotation where the codes overlap and examining the quality of the relationship Ruling out false negatives is harder and requires more cognitive flexibility It often involves re reading or search ing through text in a less directed way If there is a pattern in the data that emerges but is not found in an exist ing code or codes create a new code or codes so can reference it easily If can convince myself and hopefully the reader that a substantive relationship 1 exists and 2 refers to something real in the social world include it in my analysis and say that the failure of my initial cod ing scheme to capture this is the result of a type II error Since the substantive claim is not based on a probability of error but on an observed and demonstrable pattern or occurrence this is consistent with the logic of inference espoused above this is bracketing the issue of statisti cal induction or inference from sample to population It is a strength of qualit
25. operationalized is ultimately chosen by the researcher In essence all coding is an analytical imputation It is saying this piece of data 1 should be considered understood as a piece of data and 2 is somehow associated with the code and its larger categories Even when code names are generated using respondent speech behavior as in invivo coding there is an analytical moment whereby the researcher designates a portion of the raw data as something to be analyzed and denotes the speech behavior itself as the code with which it is tagged Before CAQDAS software codes were often simply keywords written in the margin of interview transcripts or fieldnotes newspapers pamphlets maps etc At the most basic level coding assists the analyst in understanding the massive amount of qualitative data by referencing and cross referencing key observations There are two common ways of generating codes in qualitative social science research The first is the induc tive or emergent generation of codes Most qualitative methods such as participant observation content analy sis and in depth interviewing typically have inductive components That is to say emerging previously unfore seen patterns occurrences in the data form a part of the analysis Inductive codes generated after the analysis or fieldwork has begun are used to tag and or group these events quotes into meaningful categories These codes are often then grouped into even lar
26. oth my own research and that of others have developed some thoughts and opinions that may be of use to others Dealing With and Minimizing Error My general opinion is that type errors false positives are a lot easier to rule out in qualitative research than in quantitative research A qualitative researcher can always go back and investigate a potential relationship by reread ing interviews etc looking at a text listening to audio then reassessing if the relationship is spurious based on a re analysis of the initial content In primarily quantitative research false positives are harder to deal with since the assessment is typically made by examining the numeric strength of association between measures that are largely set at the outset of a project Some recoding is possible e g collapsing an interval variable into an ordinal vari able or building up binary variables into an index but the initial categories involved in data collection and analy sis are relatively set Without additional data e g a record of the observed event it is hard to rule out the possibil ity of miscoding or mis imputation Qualitative methods furnish additional information for ruling out false positives due to errors in the coding scheme a deficient initial ana lytical scheme or chance overlap They do so by provid ing the analyst to look back at the data on which a coding association is based e g seeing if eating ice cream has an effect on
27. rovide a set of criteria specifically addressed to study deviant action as system of meanings into two related contexts the single action itself and the whole life his tory According this point paper introduces the concept of deviant career and it explains its development and steps it also provides some psychological social points e eoe o gt e e ee o o o o gt e o o o ro o o o oo Newsletter 2011 3 December 2011 of views and contributions that scientific works offers B to think about potentialities of accounting interviews as powertul tools in clinical practice and empirical research too Specifically we discuss narrative about behaviour and life story as a resource in order to take suggestions about meanings goals and rules of action It also suggests something about functions that crime discharges in the whole pathway of one s life Juridical case analysis run by means of the software ATLAS ti is aimed to demon strate theoretical reasoning proposed Please feel free to share with us any additional recent or not so recent publications that other users may find helpful or enlightening Send information to newsletter_feedback atlasti com atlas ti New Book Coming Soon Qualitative Data Analysis with ATLAS ti Susanne Friese Dr Susanne Friese Qualitative Data Analy sis with ATLAS ti Sage Publications December 2011 Qualitative Dat
28. swimming prowess and examine whether that apparent relationship is an artifact of methodological deci sions or mis imputation Ultimately this assessment comes back to the researcher and their scholarly community but in the case of qualitative research the raw data provides an additional tool for adjudication Type II errors false negatives are harder to deal with because they are invisible Broadly this is an issue of missing something either in the observation of a social phenomenon or its analysis Returning to the issue of code associations type II errors refer to real relationships that don t come up as an association in queries or output In statistical methods type II errors are often cast as less dangerous since no association is seen as the more con servative result But in reality they are still errors The analyst misses part of what is going on in the data set and the world it is meant to represent Unfortunately type II errors are harder to rule out because the analyst may not even know they are there The ability to generate codes inductively and reanalyze data in qualitative research assists in this capacity again by allowing the analyst to return to pre coded data and reformulate categories but it does not remove the problem My strategies for minimizing error in qualitative data analysis rely on the fact that it is easier to rule out false positives than false negatives in qualitative research try t
29. t is necessary to articulate what social scien tists in the positive behavioral tradition typically under stand as codes and coding in qualitative research To put it generally coding is the process by which an investigator or team marks specific pieces of data text images audio clips video geo spatial locations as instances of particu lar categories events concepts or variables In ATLAS ti codes break documents into a smaller unit of data known as a quotation The code is an indicator that a piece of data the quotation contains characteristics that are part of some larger grouping Codes can range from seem ingly straightforward demographic constructs e g age 20 29 state california to abstract theoretical notions such as cultural capital anomie or collective ef fervescence to the mezzo level constructs in between e g social network types attitudes various cultural tropes etc ATLAS ti offers numerous ways to arrange these codes into larger hierarchical and non hierarchical groupings and to search for quotations where codes or combinations of codes co occur or overlap Codes are flexible and non exclusive Many codes can be applied to one bit of data or codes can be created that apply to no data at all By applying a code to a piece of data the analyst is essentially saying this bit of data is an instance of something How this something is measured un derstood generated or
30. u know about publications of potential interest to our users that have come to our attention MihlImeyer Mentzel Agnes 2011 Das Datenkonzept von ATLAS ti und sein Gewinn f r Grounded Theory Forschungsarbeiten 74 Abs tze Forum Qualitative Sozialforschung Forum Qualitative Social Research 12 1 Art 32 http nbn resolving de urn nbn de 0114 fqs1101325 full article in German only Qualitative data analysis software is quite similar re garding the fundamental steps and procedures of data analysis but the way in which the results of the analysis are stored is different This article focuses on the fit of grounded theory methodology GTM and a network like storage as is used in ATLAS ti Using data from an ongoing research study as an example a description is provided of how the analytical steps of GTM are imple mented in ATLAS ti and how results are stored as a net of nodes This code net represents the developed grounded theory as structured conceptualized information Export ed as an XML file the code net can be used in other ap plications by other researchers This has the potential to support data quality and scientific cooperation between qualitative researchers Pan Steve Chon Kaye and Song Haiyan 2008 Visual izing Tourism Trends A Combination of ATLAS ti and BiPlot Journal of Travel Research February 2008 vol 46 no 3 339 348 By analyzing 766 articles from travel trade magazines pub
Download Pdf Manuals
Related Search