Content Analysis

hi this is shady Artie a professor at leas University in Belgium and today I'm going to present you content analysis content Inessa analysis is a method under qualitative research methodologies this presentation is part of a playlist on qualitative research methods I would like to advise you to see the previous presentation mainly a presentation called qualitative research method well when we look at the different APIs Thoma logical or research methods or research approaches in the scientific world we can group research under two categories quantitative research methods involving modeling empirical and experimental approaches and qualitative methodologies and today under the qualitative methodology I'm going to focus on content analysis previously I explained what is qualitative research what is systematic review and here today we are focusing on content analysis well don't forget that some researchers they select to combine qualitative with qualitative and distance we call it the mixed research methods well when we look at qualitative research methods there are two types of data collection mainly one is focused on people and one is focused on document when we talk about content analysis it's mainly focused on documents well therefore the presentation audience is mainly researchers who are preparing our research thesis or a dissertation or a conference or journal publication in relation to content analysis the aim of this presentation is to enhance your capacity to conceptualize design and conduct content analysis based research I'm going to explore the definition of content analysis what are the methodological approaches how to do data processing how to validate our content analysis and finally some lessons learned lessons let's start with content analysis what it is simply content analysis determines the presence of certain themes or concept within some given qualitative data it can be text it can be images to quantify and analyze the presence meanings and relationships of such themes or concepts and in general content analysis is considered one of the desk research what are the types of rescue desk research first of all systematic reviews secondly content analysis and historical analysis those are types of desk researches which are considered as research methods under the qualitative research methodology well content analysis is simply looking at transforming a large amount of text or images into a highly organized and concise conceptual structure we try here to have a systematic process to ensure objective analysis and we do qualitative data treatment for text or images and we start with looking at a process called coding coding process is simply to organize communication content whether it's textual material or visual images or illustrations according to communication schemes and we look to identify content relevant to our research question in general and the findings subject our subject the findings subjected to quantitative and qualitative analysis or both they are dependent on that well now what are the types of content analysis we can use it to characterize responses in open-ended surveys questions for example focus group and interview transcription so it's not only transcripts or manuscripts that are already published or available or documents or archives or records I can also do content analysis for an interview that I content I conducted for example I can do a Content analysis for the report of a focus group discussion so simply it's a technique it's a method that allows you to go in-depth in any Content and to analyze the structure eyes and to come up out with the main concepts and structure eyes them in general and we look also to another type when we look at the evaluation of content trends and journals or magazines public records newspapers textbooks or cookbooks you can even conduct content analysis when somebody is performing an interview with you and you want to lie to highlight what was the focus and the terminology and the words and the concept the ideas that are there and here we can also start a concept analysis content analysis so content analysis depend on the purpose but it's simply that you have a content and you try to analyze it in order to structure it and make a meaning out of it in a short and brief way some examples or characteristic of content analysis when it's done in a qualitative context we look at examining meanings of the content for the development of text images theories based on researchers knowledge and evidence drawn from the study and the process involves identifying the relevant data coding and generating themes based on the underlying meaning of the data and it involves using themes to address research questions so this is part of the content analysis now what are the major methodological approach of content analysis well in general we have to go in a certain process we start to choose the data sources that we will work on code this data develop some categories assess the validity of the reliability and finally start to analyze our results so these are the major steps when we do content analysis and we can have a certain approach when we do a Content analysis we can have a different approach either it's inductive so we try to examine without any preconceived notion or category and here we are not having any theory in background we just have the data and we are looking at interpreting it we want to explore generally a relationship or we can follow a deductive approach where we are having a predetermined keywords and categories and variables and we want to check our day present in this text or not and this what we call is the deductive approach in general when we talk about the methodologies to do for content analysis either we do we have to have two main access the first of all intent we have to determine the analysis purpose and the type of outcome data that is desired do we do here inductive or deductive approach are we examining the manifest of obvious or hidden content do we look at deeper meanings that are hidden before behind the terminology or we're just taking the words literally as they are what are the methods of validity what are the methods of reliability so this is the first thing when we start to content analysis we look at the intent in January and the second part we look at the technology so how are we going to do that which is mainly through computerized coding so we start to code the data and try to put it in a computer to make sure we can have analysis now how will the content analysis content will be examined there is two approaches either we focus on the manifest so here we look only on visible meaning at the server level of the text and here we are just looking at coding and keyword searches and some people mainly psychology people working in the cycle or psychology field they are looking at the latent approach while we are looking at deeper meaning implied in the text and here you need to have an expertise in this field in order to conduct more interesting and debatable interpretation of what is behind what was said what was behind what was found in the content that I am analyzing now how to do the data processing for content analysis today in our modern word we do it through certain steps we get familiar first with the data we revisit the research objectives to make sure that we have a clear question so that we do not get lost in the process we identify the patterns and the connections and then we start to develop a framework that fantasizes or summarizes our understanding of our analysis so one of the major steps that we need to learn when we do content analysis is to do coding and classification what is coding and what is classification well coding is simply segmenting we start to divide the data into meaningful analytical units and those meaningful analytical units we start to code them so we make segments of data with symbols descriptive words or category names and we start to group these codes under certain categories so keep the continued ality develop master list during the coding and we start to read the text and we develop the codes that corresponds to the concept the main themes that are addressed or that are found in the text and we start to code them according to common our recurrence in the text so this is the first step when we do the coding well what is coding in general there is different types of coding either we do inductive coding so developed by researcher by directly examining the data all we do a priori codes so they are brought to the research study or developed before examining the data so we have already a code with our assumptions all we can do co-occurring codes so partially or completely overlapping codes same lines or segments that can have more than one code so these are the different approaches when we do coding and for sure in deductive studies codes might have been predefined whereas in inductive code studies coding is vital to look for the unexpected so this is the the main difference between inductive and deductive approaches when we come to coding and in general I use the inductive coding method because I try to without pre assumptions try to interpret the text and come up with interpretation or a summary of it this is an example for a coding process that is used in a certain in a software and as you can see here these are this is the text so when you have the content you can upload it in a software that treats the information so this is the text once you have a sentence or a term or a word or a concept you start to highlight it you highlighted you directly connected to a specific code that you developed and created that relates this theme or this text to the code and simply the software can localize this code connected to the document and directly link it to the location and from there you start to code your whole text every time you find a term on your list you have a list of codes that you are already developed every time you find somebody or something written on this code you start to identify and later on the software will count how many times this code was mentioned for example here the term maintenance was found 46 times in the docket and I can click on the word maintenance and I find all the quotations where it was found so this is the way how to structure my content and simply coding software's help you to code and help you to draw a framework later on to group these codes under categories that are meaningful and finally you come up with a framework of understanding so when you select the main codes for your study you have to group them under categories and for sure you will find out that not all codes are important so you will need to eliminate some codes that are not relevant to the study or do not bring much in relation to your research question and then you start to categorize them and these are the most important that this is the most important step that you need to do moving from a code to categories so this is a higher level of understanding and this is a high level of analysis well how do I do the classification and how can I move from coding to categorization you need to count the number of times a code was mentioned you need to categorize the codes into primary and secondary categories to make sure what is important was less important you need to recode and choose the best fitting code names for a cluster of topic among concepts in relation to a research question and then you can come up with your framework of understanding that summarize all this large content into a meaningful interpretation well once we are done with the categorization so we coded we categorized we go now building our conceptual framework a conceptual framework is done the same approach not only in content analysis we do the same approach when we do content analysis for interviews or group focus group discussion we start by coding and from the coding we come up with a theory or with an understanding and we create a structured version on the topic and insights on the topics so this is the main purpose of content analysis well how can we do that for sure there is manual approaches and there is computerized approaches I advise you to use computerized approaches but for sure it depends on many factors the amount of material to be analyzed it depends on the number of researcher involved and the level of experience it depends on the financial constraints and the long term goals and it depends on the availability of electrical electrical electronical devices in your context where you have word scanners that you can scan if you have a word recognition software or not and these are all factors because these software's that I showed you some of them is are not fulfill most of them are paid software so you make sure that you have access to them and you have to look at the comfort level while conducting your analysis well this is an example for some examples working from paper-based approaches of content analysis whiteboard sticky notes Excel access and more advanced sophisticated software previously people were doing sticky notes when they do content analysis so they will create maybe a workshop or they will group people they will all get into the document they start to analyze it they code the text into sticky notes they hang it on the wall and I start to group the different concept and they do the whole coding categorization and framework creation in sticky note approach some others do it with papers also a lot of researchers do use Excel and in distance they try to group the concept and they try to do the analysis because they find it a software that is easy to access and they start to group the codes and they start to link them to the text however it is tedious and my advice is simply is to focus on professional software's that help you with the content analysis so the first step always is to have initial consideration when you start your process identify the data that needs to be analyzed and you look at the sources of the data and you start to look at the words and the codes then you start the creation of your coding this is step two step through three is to work on the coding data and the key words and the key concepts that you have you find them and you start to locate them and you start to highlight them and connect them then you check the reliability of your approach you compare your codes you look at the resolution of differences you can get appear or another I of another expert looking at your coding approach and then finally you start to analyze your result you look at examples you count number of codes you look at any cross tabs or relations and you start to look at any incorporation of qualitative data related to your coding so these are the general steps that you will need to address and doing that through a software is much much easier however if you prefer to do it through Excel or use sticky notes or even a whiteboard or paper based approach it will just be more tedious now I'm done with my analysis and I need to validate my content analysis as I told you before I always say qualitative research methods are by default subjective and I need to prove that I am NOT biased so how can I avoid bias this is a common slide I always show talking about external validity and the internal validity do the qualitative researchers think about the reliability and the validity in the same way in quantitative researchers yes in qualitative research we always look at validity similar to the quantitative research methodologies but how can we do that how can we avoid bias we try to look at the credibility and confirm ability look at intention and relevé reliability did we code appropriately did we select write terms do they make sense can we have other experts sharing their vision and share revise our coding process this is for example an example for internal validity and finally we go also for external validity can I generalize my outcome can my findings be transferable or are they dependent on other factors so these are all approaches and when we do validity we look at when when selecting the content to be studies when the sample selection methods that we are working on so we need to first when we do content analysis to look at that what is the content that I have is it representative or not we need to construct the validity by using previous variables categories and then testing new ones and the coding system so we can try to revise our coding system and our categorization and we check does it make sense or not is the one method or no our proposed structure is bringing the best out of the analysis of the content or not so these are all approaches and we need I'm sure that the research question is answered so this is very important point also as part of our validity for content analysis we can use the interquartile approach what is the in decoder it's an independent coder that can analyze the same text using the same categorizing coding scheme and reach the same decisions and this brings reliability and consistent consistency so this is obviously part of the internal validity I can increase the internal validity of the work by inviting an inter coder exactly like the fight another approach that I can do I can do something called the intra coder and in this sense I am looking to the reliability within a single coder and I asked to recode a subset of data by the same person again to make sure that this person is consistent in his or her coding system and Here I am looking at ensuring coding decisions where that are not altered so when we look at internal reliability in a Content analysis we invite inter coder and intra coder and with those two approaches we can increase the quality and make sure that we have a more rigorous approach in coding and we are avoid as much as possible bias well I'm going to terminate or end my presentation and I would like to share with you some final learned lessons when we do content analysis first of all keep the research question always nearby so that it guides you so that you do not get lost because content analysis is one of those studies that are time consuming and you can easily draft drift or fade away while you are doing it so always keep the research question under your eye content analysis is time consuming and it depends on your case so this is very important to know that when you are investing in content analysis it will be time consuming and therefore we need to use software's that can help us and shorten this time an especially software that can do text recognition especially with new artificial intelligence software this can be a potential to make it a shorter process and if software is used to assist for the categorization and organization of data the researcher must create a highly logical coding category and categorization framework and I had voice you always look at validity the enter code and int recorder are two approaches to make sure and sure and assure the quality and make sure that you are avoiding bias and make sure that you have a higher internal validity of your account and outcomes and as usual always learn from other journal papers look at a journal paper with a title content analysis read it well and if it's well cited most probably than they followed a good methodology and then you can imitate the methodology especially when you are doing it for the first time so by that I would like to end up my presentation content analysis is one of the desktop methods it's one of the qualitative research methods that are used in qualitative research I wish you rest with I wish you the best with your research thank you for your attention

Channel: Shady Attia

Views: 34,362

Rating: 4.9134469 out of 5

Keywords:

Id: sOt1FOWKGKA

Length: 20min 43sec (1243 seconds)

Published: Sun Mar 01 2020