In Unit 2, we're going to discuss theory, conceptual frameworks, theoretical frameworks. We're going to talk about definitions of variables and specifically divide them into operational definitions and conceptual definitions. We're going to look at how variables are measured and different levels of measurement. And we're going to begin our qualitative unit discovery in this section as well.
So research either tests or develops theory. And remember, we're going to base everything back on those three types of questions. So if we have an explore type of question, a compare or relationship type of question, or an intervention or hypothesis question, we're going to have different levels of theory if we select those type of questions. So for example, in an exploratory question, there may be literature on a population or on the topic you're interested in, but there isn't a lot of research on both.
For example, we might know a lot about falls, as well as measures to decrease falls in that hospitalized patient setting. But we don't know a lot about some populations. For example, moms post-delivery and falls, or postpartum falls. And there's a little bit of research on falling babies or dropped babies.
There's not a whole lot of research about moms falling postpartum. So if you're interested in looking at that research question, You look in the literature and you don't find a lot of literature there, you're going to be looking at more of an explorer type of question or a relationship type question. You're probably not going to be looking at interventions.
So again we're going to move more towards exploratory or descriptive study. And this exploratory or descriptive study is going to be one that helps build theory. So as we gain knowledge we can build theory with those discoveries. As opposed to if you're interested in preventing falls in the inpatient setting, perhaps in an orthopedic unit, there's a lot of information on that. So you're going to have a lot of existing knowledge, a lot of existing theory, so you might move more towards an intervention-type study.
When we look at the comparative types of questions, relationship questions, or comparing two variables, two or more variables, There should be some evidence that suggests to us that a relationship might exist between the two variables you're interested in. So there must be some type of theory that leads us to believe two things are related and we might be able to add or inform to existing theory based on the findings or results from that study. So an example is there, back to our falls example, there are some research Studies that look at the relationship of propriocessence and falls in the elderly and this is in the physiotherapy research and these have led to certain interventions related to gait training to prevent falls in the elderly.
So we're still interested in falls postpartum but if we can connect that variable of change in proprioception as something that exists in the postpartum patient we might be able to see a connection between those interventions used for elderly and proprioception and the interventions used if we were to use them in postpartum moms. So we first would need a study looking at proprioception and postpartum moms. That would be the compare relationship type of question.
Once that's established and you have evidence, then you know a little bit more about your subject and then you might actually go to an intervention study where you're actually using some of those interventions or gait training interventions to see if they are effective in the postpartum population. So you can kind of see how we've built from exploratory to compare relationship to intervention types of research. So when a researcher says they are a certain type of researcher or another type of researcher, say qual and quantitative, really it should be your question.
What do you want to learn? What do you want to discover that should lead the type of question? So what is the knowledge? What knowledge is available in the area that you're interested in? So we are going to come back to that, but we're going to move on to focus a little bit on theory.
Sorry, so we've been discussing the relationship between theory and those three types of questions. So we're going to move on now to discuss a little bit more different parts of a theory and a little bit more about... Theory.
So concept here, we have missed that, but a theory is made up of a description of the relationship between certain concepts. And a concept is a description of a class of phenomenon. It's a single abstract idea.
So it includes things like addiction. Addiction is a concept. Hopelessness is a concept.
Nausea, stress. So what happens with a theory is that we try to explain the relationship between individual concepts. Now a model is a representation that bears some degree of similarity to aspects of the real world it represents. Here you see a model used in art classes commonly to represent the human body, how different shadowing might, how different shadows are cast on the body.
I can help beginning Artists and actually advanced artists in representation of their art form if it is a human body. So a model in research, again, is going to have some degree of simulator to aspects of the real world it represents. One example from nursing is the health belief model that we use quite a bit.
The health belief model. is represented here. You can see the different concepts of perceived threat, the expectation or outcome expectations, perceived benefit, perceived barriers, perceived seriousness of consequences of the problem, perceived susceptibility to a problem.
So these two things, the perceived susceptibility to the problem and the perceived seriousness of consequences of the problem, relate to perception threat, which relates to self- efficacy or the patient's perceived ability to carry out recommended action. And then on the bottom half of the model you see perceived benefit of their action or perceived barriers to taking action affect outcome expectations which affects self-efficacy. So that's an example of one model that we use and this model can also be used in a study that someone is setting up. You can base your study on it by connecting the variables of your study to these concepts.
So if we were looking at perceived susceptibility to a problem, so that might be perception of asthma symptoms or seriousness of the consequence of the problem. risk of hospitalization or you know however you are gonna in your study what things in your study represent these concepts so you would connect that those things together and that's the way that a model or existing theory can help support or direct research and there are numerous examples in nursing the self-care model Lots of different models that you can look at. Now, when I was saying you could use an existing model and link that up to your study, when we do that we call that a conceptual model or framework. And that's a network of concepts in relationship and they account for a broad nursing phenomenon.
An example of that is the self-care model and we can use that as the basis for study, research studies. Now it gets a little confusing because the actual conceptual model or framework might be published as that, just the model itself, like the self-care model, but then also within the study you have a conceptual model or framework that directs the study and that can either be an existing model or framework, so one that's published like the self-care model or health belief model, or you can take existing research findings and Use that as a knowledge base to create your conceptual model or framework for your study. So there's slight differences there. So you might see that term used, and it can be somewhat confusing sometimes. So let's move on to sort of general theory.
General theory takes all instances of nursing into consideration. Remember, nursing is concerned with three concepts, person, environment, and health. So examples. of this general theory, or sometimes it's called grand theory, includes a more global look at persons, environment, and health.
So here we have Sister Cluster Roy and the Roy adaptation model. So again it's confusing because the word model is used and we just talked about conceptual and theoretical models but this is actually a grand theory. So it is confusing but this theory does attempt to explain all of nursing, the relationship between person environment and health.
And if you're coming from within this grand theory, you are going to design and execute your study differently. If you're coming from a adaptation model, your methods may be more linear, quantitative, as opposed to looking at Watson's holistic nursing or caring model. That could be either, but tends to be less deterministic, less linear. And Newman, Health is Expanding Consciousness, is always, almost, well I would say always, a qualitative inquiry.
And Newman's model looks at health patterns or patterns of the person. And within that grand theory, there really is no better or worse health. It's just health as your current state.
and our health as an expanding consciousness so they really don't quantify good or bad within that within that model so there are no higher levels of health or things like this so anyway the purpose of our class is not to go in too much to nursing theory I know you've had that but I do want you to know that research should be based on theory and or inform theory And then that's the way that we build knowledge within our discipline. And then we have middle range theories. These represent a partial view of practice.
One example from my nurse practitioner days was Cox's model of primary care. So that is a middle range theory and it's only going to apply to nurse practitioners in primary care. those concepts that would predict health or healthy behaviors. And then we have micro range theories and this is Afaf-Maliz here, the situation specific theory. And this one is really, I love this one because it basically is given the particular situation, you are going to use knowledge best suited for that particular situation.
And it reminds me very much of how I approach different patients in practice. So I kind of like that with research. Because if I'm going into a room, how I might interact with a patient is going to be based on more on the characteristics of the patient than the characteristics of myself as the nurse.
And that's just how I practice nursing. Let's see. So conceptual models or framework or theoretical frameworks.
Again, I discussed how... This can be used in two different ways. So we're going to discuss this as concepts and or theories pulled together as a map for the study.
Again, an existing theory is often used, but each variable in the study must be matched to the concept in the model, or at least a portion of the model. Every study doesn't have to test an entire model. You can test just a section or different parts. Like in that health belief model we just discussed, we might just look at... outcome expectations.
We might not be interested in perceived threat in our study so we might just focus on that part of the model. You really need to evaluate that it is an appropriate connection, that somebody hasn't just matched up two things and said this represents perceived barriers to taking action. It should make sense and as a reader of research you can look at it and decide whether you think it makes sense. So and also it's going to determine a lot about how how your design of your study evolves. So if your framework comes from a biomedical model, then your measures will be biomedical.
If it's psychosocial, the measures will be psychosocial. So when I say biomedical, you might be measuring blood pressure, weight, might be collecting blood, things like this. If it's psychosocial, you might be doing more interview or doing surveys or self-report.
So really your model is going to kind of determine a lot about how you conduct the study. Everything should be consistent and again all falls back to the original. ...question. So you really, you can't, I can't overstate how important it is to take time and pick and write the actual question carefully so that everything kind of flows from that because it all begins with a very well-developed question.
Theoretical frameworks is a structure of concepts that exist in the literature that's a ready made-up map. that you must make it clear to the reader how the theoretical framework connects to the variables in your study. So some texts will break this apart, and as I said it is hard because people use these terms interchangeably, but theoretical framework we can, for the purposes of this course, will be an existing framework in the literature.
A conceptual model or conceptual framework would be connecting one from the literature to the study or connecting literature in a conceptual model for the study. Okay, so what? So at this point you're probably struggling and that's okay. This is very normal for most students. The trouble is that we're using language and terms that have meaning already in our everyday speak.
But now in research terms they're using in other contexts that have specific meaning to research. So my best advice at this point is that you treat this as a vocabulary test and learn all the terms described in your readings and within these recorded units and it will make it much easier for you to identify the various parts of a research article as we move along. If you're still struggling with the term what it is you will have trouble finding it because most of the exercises the classroom assessments and exams are based on your ability to apply this these concepts. So we must move past to knowledge, a knowledge based or knowledge kind of recall to application of this knowledge.
So with that in mind, we're going to move on to one of my favorite sections. Literature teaches us about variables of study. So, you know, you have to read widely.
If you're interested in a topic, you're going to read and see what's known in the literature about that topic. And you need to develop precision of terms so that when you describe a variable there's no question as to what you're describing and you need to decide how to define this variable for your study and you'll need two distinct definitions for each variable of study. One is what the variable is and the other is how are you measuring it. And this sounds easy but it becomes very important and it has to be consistent with the type of question that you're that you're asking.
Let's look at what it is, is a conceptual definition. How you measure it in the study is an operational definition. So let's look at anxiety as an example. So anxiety is the variable and let's look at some conceptual definitions for that. So Anxiety, a vague feelings of alarm that persons report when faced with a stressful situation.
So you can see from that example that in order to be consistent with this conceptual definition of anxiety, your operational definition, or how you're measuring anxiety, should be self-report. Because it's the vague feeling of alarm that persons report. So persons report.
So that's your conceptual definition. So your operational definition is probably going to be some measure of self-report. Let's try another conceptual definition for anxiety.
A trait possessed by all persons to some degree which is reflected in their responses to questions and infers anxiety level from those responses. In order to be consistent, an operational definition for this conceptual definition of anxiety would include an interview or open-ended item on a survey, perhaps. Let's try another. Behavioral manifestations of persons subjected to stress which can be identified by grimaces, muscle tensing, and palmar sweating.
Given this conceptual definition, an operational definition must follow to include observer's perception of behavior. This might include measures of muscle tense, palmar sweat, other ways to measure grimaces. There's not just one way to measure it, so it's important that your operational definitions are written such that there can be no misunderstanding as to how you will be measuring that variable in the study.
So characteristics of a good operational deposition is that it's consistent with your conceptual framework. Has logical and empirical meaning and defines your concepts explicitly and precisely. So precision.
So let's go back again to question type. Definitions are less precise in exploratory or descriptive question types. If the point of the study is to find out about those variables, an example is from my own work on positive and negative dating experiences. I didn't want a specific definition for dating because it was an exploratory study. Instead, my definition of dating was participant-defined.
But definitions will be more precise in intervention types of study. So if you're looking at, like, weight, If you're, you might define weight as, is that on the next, nope, you know, you might define weight as, I don't know, the force of gravity on, I think we talked about that in the last lecture, whatever the conceptual definition for weight is. And, but if we operationalize that in an intervention study, it's going to be very specific.
So weight in. kilograms on a standing scale first thing in the morning with whatever types of clothing would be specified as the operational definition of weight. Now in another study that is not as precise you might have weight measured as more of a subjective measure. So you might say within normal weight, overweight, obese or overweight. People might be checking the boxes, doing their own self-assessment.
So you can see the same variable weight can be measured in very precise ways or not so precise ways. So even more precise ways is when they actually put you in a tank of water and do a weight that way. You might even be looking at different percent fat and that kind of thing.
So different levels of precision for different types of studies and that should make sense. So if you see an exploratory study with a very precise definition of the variable that might not make sense. You have to be reading all of this very critically.
So let's look at more at types of variables. So independent variable is the influence. So it changes, a change in the independent variable will be measured in the dependent variable. It's not necessarily the cause, but the independent variable, if it is an intervention study, is the one that is manipulated.
So one person gets the independent variable and then the other person perhaps doesn't, and then you measure the dependent variable and there should be a difference. The dependent variable is dependent on the independent variable and is the one where the change is measured, as I said. The independent variable is manipulated, the dependent variable is not. And there can be more than one independent and dependent variable, depending on the study.
So you have to read carefully. And then not all studies have an independent and dependent variable. The study I just showed you just had variables we were exploring. There was no independent or dependent. And a comparative question or relationship question might say there's a relationship.
But neither variable is independent or dependent, or sometimes it can be inferred based on the question. So what is the relationship between smoking and lung disease? In this case, the independent variable is inferred to be smoking. The dependent variable is the measured change in lung disease. Intervention studies always have an independent and dependent variable.
So let's go back to our questions again. Exploratory questions do not have an independent and dependent variable. They just have variables.
Relationship type questions may or may not. And an intervention study always has an independent and dependent variable because you have an intervention. And again, you might have just two, there's always at least two in an intervention study, independent and dependent.
There's always two variables in a compare study. In an exploratory study, there might just be one. Usually it's just one.
So you really look to your theoretical or conceptual framework for clues if you're unsure what type of variables are in studies that you're reviewing. Because remember everything should fit together like a puzzle. And we have other variables as well to consider and you should know all these terms. Intervening.
These are also sometimes called confounding variables. These intervening variables directly affect the independence variable effect on the dependent variable. Again these intervening variables directly affect the independent variables effect. On the dependent variable. So look at the question.
What is the relationship between caloric intake and weight gain? That's the research question. What is the relationship?
between caloric intake and weight gain. So an intervening variable could be activity level because that's going to have a direct impact on metabolic weight. Sorry, metabolic rate, which would have an effect on weight, which is the dependent variable. Sorry.
So that's intervening. Now extraneous variable are similar but they are not of interest to the research, but they must be controlled for in the question. So in the question above, examples that might be extraneous variables to the question of weight gain might include age, family history, obesity.
So these might kind of compete with the independent variable as an explanation for change in the dependent variable. So these need to be controlled for. So how variables are measured.
A variable implies it must vary even if it's only positive or negative or present or not present. So yes or no is a variable. Categories must be mutually exclusive and the categories must be logical and consistent with theory.
So for example, people set up age categories. You can't just willy-nilly randomly put 0 to 5, 5 to 10, 10 to 15. they have to make theoretical sense. So, for example, you might put in age categories as infant, toddler, preschool, school age, adolescent, early adults, and so on.
So if you're putting in, if you're collecting not the specific age, but you're putting people into a range, then you need, it needs to make sense, the categories that you're placing people in. Married, divorced, not married, or, well, single, partner, whatever. It has to be based on...
some kind of existing theory or structure. And then there's different levels of measurement. All variables need to be defined and levels of specificity will depend on the question or the level of precision, which we just talked about. The goal in intervention studies is to be precise as possible. And again, in exploratory studies, you do want to be clear as to what you are interested in so that you're not collecting.
Lots of data about things you're not interested in, for example, anxiety and fear is a very hard one. You might think you're collecting information about anxiety when you're actually collecting information about fear. So you have to be very careful that there is a lot of confidence and the measures are specific to the concept you're interested in. You need to define what it is and what it is not and the levels of measurement will determine data collection and analysis so the researcher must be consistent.
So when we talk about level of measurement, and we're going to talk about nominal, ordinal, interval, and ratio, there are certain statistical measures that go with this, or non-statistical measures. So they need to be consistent throughout throughout the study. And sometimes you can catch a researcher because if they collected ratio level data, which is a higher level of data, and you can use more robust statistics, and suddenly they're reporting ordinal or nominal, they've transferred it to this type of data and they've changed their stats. Sometimes that's a clue that they didn't necessarily find what they wanted to find.
They've started to manipulate the data. But anyway, that's kind of fun and you can do some real critical reading to see what the researcher has done with that. So let's look at nominal. So nominal level of data is mutually exclusive that there can only be one category. There's no numeric meaning to it.
There is a spot for everyone. So examples of this could be gender, ethnicity, and marital status. That's examples of nominal level data.
It's important to know these because later we're going to talk about the different tests that go with this kind of data. Ordinal, they're scales of increasing magnitude, they're ranking, they are also mutually exclusive. But you can't perform mathematical operations with it.
So yes, sometimes no. There can be an increasing level, but they're not equidistant from each other. So my no might be a big, big, big no.
Your no might be meh, no. So you can't really perform mathematical operations at the distance between those categories. But they do have increasing magnitude or decreasing magnitude between them.
So that's ordinal. You can put them in order. So if you're trying to differentiate nominal and ordinal, nominal is discrete and exclusive, but they don't have any particular order, male, female, there's no order there. Ordinal, you can put something in order.
And then interval level, they are equidistant intervals or units. We have the, let me see if I can find this. Okay, so if we look here, here's an example of an interval, actually this one has an absolute zero, so this one we actually can look at as a ratio, but if you forget the lines at the top, say you just had the faces, you didn't have the zero per se, maybe that's not the best one. Let's look at this one.
Oh here we go. So if you look at this no pain, pain as bad as it could be. If somebody makes a mark on this line then you can measure the distances between it.
So Whereas at the top, no pain, mild pain, moderate pain, severe pain, very severe pain, worse pain, you don't know really, you can't really measure the true distance between here and here. Because very severe pain for me and worst possible pain can be very different. So in this case, you're looking more at ordinal. descriptions. Now because they put this line above it, these are equidistant, so then you're looking at interval level as well as all of these you can interpret into interval level.
And the key thing is you can also have ratio level because you have a zero. If you have an absolute zero, Then you have ratio level data. So ratio is absolute zero.
That could be measures of blood in the urine because they're not supposed to be any blood in the urine. The Calvin scale is ratio level, whereas the Fahrenheit and Celsius are interval because there's no absolute zero. Now ordinal, so all of these are interval or ratio level.
So the ones with the absolute zero become ratio. Now again, if you had these with, if you had these descriptors without these lines and without these measurements, then you would just have ordinal. And let's see if we can find one of those.
Okay, here's an example of a Likert scale. So you can see there's no way to measure the distance between, well, The way they have it set up is kind of funny, but anyway there's no absolute zero here. There's no absolute distance between strongly agree, agree, agree, neutral, disagree. Whereas the Payne scale actually had numbers 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. So you can see the distance between 2 and 4 is an actual number, so it has mathematical meaning.
The difference between strongly agree and agree, there's no mathematical meaning there. So This example here would be ordinal. So you want to be able to get into the habit of being able to decide what you know what is it nominal, is it ordinal, is it interval level data. And the confusion often comes just make sure that you know there has to be a distance between each of the options and it can't be subjective distance. So I think we've given some examples there.
You might also look at like blood pressure. So blood pressure could be interval but it doesn't have an absolute zero right because if you had a zero blood pressure that would be dead. So blood pressure would be interval level data. Okay so we're going to come back to those throughout.
All of this is sort of the foundation blocks and we keep returning to these terms. So it's really important to kind of jot down and learn these terms early in the course and you'll have less trouble later. So now we're going to move on to qualitative methods. This is one of my favorite areas of research. Most of my own publications are qualitative.
...educations with some survey research. A lot of my work in my graduate studies was on quantitative work as a graduate assistant, entering data. So I have some experience in all areas, but this is my favorite type to do.
And you are going to have a little practice at it as well with the observing an animal assignment. So qualitative research methods... The essence of it is to ask a question about human experience. Qualitative research is often conducted in natural settings, within the setting that the experience is occurring, and uses data that are words or text rather than numerical data.
You shouldn't see numbers in qualitative research. Now that isn't to say that they don't describe the number of people they interviewed. perhaps their income, you know, they'll collect those type of statistics, descriptive statistics on their study participants, but when it gets to data analysis they're not using numbers, we're using text or words and description and finding meaning from those text words or description.
You'll also see, and I'll show you an example today about some people that use photography in qualitative and analysis is really kind of fascinating and very useful tool. So they're using these words, texts, interviews, observations to describe experiences that are being studied. It's really important that you understand the distinction between qualitative and quantitative because a lot of these activities I've described you can also do quantitative research that also uses observation but they quantify it.
They observe something and quantify it. and then use that quantified number in their research. Whereas a qualitative researcher is doing observations and the result is a narrative description of what they're observing. So it's different.
So you want to read carefully. I find a very common mistake for students is they'll find a quantitative numbers research study that is descriptive or exploratory. Remember those levels of questions?
You can have a quantitative, descriptive, exploratory study that just describes something. a population or what have you and describes it in numbers. A qualitative exploratory descriptive is going to describe in narrative text. Okay, so try to see if you can make that distinction in the research articles that you're reading. So let's go through to the different types of...
One other thing I want to point out is qualitative research is more inductive. from unit one when we discussed that, whereas quantitative research tends to be more deductive. Okay, so first we're going to look at different types.
So ethnography is a description or interpretation of culture, and the goal is to understand the emic or insider view as opposed to the etic, outsider view. What you have, particularly if the researcher is an outsider, is a combination of this insider-outsider view. This research approach was designed to produce cultural theory, and the ethnographic method scientifically describes cultural groups. And the goal of ethnography is to understand the emic or native view of the world.
Cognitive models or patterns of behavior within a culture again are going to combine this emic-etic view together. The essence of the method is that results in a description of a cultural group or subgroup. The foundation of ethnography is in cultural anthropology. Remember I told you that a lot of nursing scholars got their start in getting PhDs in programs outside of nursing because we didn't have PhDs at the time. And you can see this in our research where we are often drawing from other disciplines and the methods that we use.
So here's one example here. It's a grounded theory of female adolescents dating experiences and factors. Phenomenology is an approach aimed to describe experience as it is lived in order to understand the meaning of that experience for those that have it.
Okay, so let's go back to case study. I think, oh, was that in there? Oh, we skipped historical.
So historical research is a little bit different. The systematic compilation of data and the critical presentation, evaluation, and interpretation of facts regarding people, events, and occurrences of the past. So using historical methodology might shed light on past and in the future.
Different sources that might be used is an eyewitness account as a primary source. A secondary source could be a view of the phenomenon or another perspective, another's perspective. This method involves some external criticism of the documents used.
So is the document real if they're using documents? Internal criticism. You look at the reliability of the information in the document. So an example might be you find the journal of a patient who has schizophrenia and yes, external criticism of the document is real, internal criticism is this a good historical record? Maybe not.
So the essence of this method is systematic compilation of data to describe some past event and the foundation is in philosophy, art, and science. So case study is the study of selective phenomenon that provides an in-depth description ...of its dimension and processes. The case study method investigates a contemporary phenomenon over time to provide an in-depth description of essential dimensions and processes of that phenomenon. Unstructured information from several sources might be used.
An example might be the use of Ambien in brain injury. Sort of a mother one night, her son wasn't sleeping well and she put one of her Ambien in his g-tube. and he became more awake and was talking. So anyway, if you're interested in that, anyway, that came about with case study research, and there's still case study research going on with that because it's only a small population that this actually works for, and the doses are pretty high.
But if you Google Ambien and brain injury, you'll see some interesting videos on that. But anyway... The case study might be used in rare disorders, rare situations, something we don't know a lot about.
There are some problems with being able to generalize those findings, but it is a good way to study like a new infectious disease or again a new reaction to a drug that we don't we don't know about, maybe therapeutic or non-therapeutic. We also, Newman, the grand Theorists that I discussed earlier, you could describe the pattern analysis as a case study form of research. So levels of evidence within qualitative forms of research, grounded theory is considered more robust than say phenomenology and ethnography is kind of in the middle. So if you're thinking about how robust or how to use this information, again qualitative is more naturalistic, quantitative. relies more on positivism, empirical and analytical thinking.
So let's look at sampling. So in qualitative design, purposive sampling is you select the participant or the observation or whatever you're you're doing in that qualitative study for specific purpose or experience. So I selected women who were pregnant post in vitro fertilization because they were pregnant post-vitro fertilization. I didn't select patients who were pregnant any other way.
or patients who weren't successful, because I was interested in that specific thing. So it's proposive sampling. Judgmental sampling, some of these words mean the same thing, but you might see these terms.
Maximum variation, you might want to, you know, if I was looking at dating violence, I sort of wanted to look at all of dating, so I really did want maximum variation. I wanted to look at all different types of teens at first. So I would select teens for that reason.
Extreme case in a study I did later I wanted to interview women who had high scores on violence measures so I first had them fill out a danger assessment screen and an abuse assessment scale and only those who scored very high were included in the study because I wanted to describe the lived experience of extreme cases of abuse in adolescents. Typical case that would refer back to my dating study where I just wanted the typical teenager and dating. Opportunistic sampling is used a lot in qualitative research. Snowball sampling is when, snowball networking and chain are all kind of the same thing, where you meet somebody and they know somebody else within the group and refer that person to you and then you interview that person.
This works really well when you're looking at rare diseases like Rare genetic disorders because, or children with certain learning disabilities, because any group that tends to organize or collect... or socialize together. Military groups is another example how you might use that snowball or network because once you get one participant it's easier to get more because that participant knows a bit about you and then also has access to other persons similar to them and that that similarity is of your research interest that's very helpful.
So the other thing that's different in qualitative is sample size. You do not, depending on what type of study it is, you need more or less. Case study, you just need one.
Phenomenology could be 10 or fewer. Grounded theory, you tend to need more to develop a theory, so more like 20 or so. But if you see qualitative studies with subjects of 50, you know, these really high numbers, that doesn't, that isn't necessarily a strength.
The ability to really compare and really evaluate the narrative would be very difficult in large studies. So you're getting more than 25 or so. It would be very difficult to be really immersing yourself in the data and immersing yourself in analysis. But how we decide when we're done is data saturation. So it's not that we pick ahead of time, oh.
I need 10 subjects or I need this many subjects. You keep interviewing, keep collecting until you have heard all the themes before and no new themes are emerging. So how comprehensively and completely the research questions were answered.
That's data saturation or theoretical saturation it's sometimes called. Assessing if the data collection procedure is appropriate to the method. So in qualitative research you do want to see does it make sense.
Did they use observation or field observation? Did they use interviews? Did they review documents?
So I gave you the example of that photo novella with the teens. That's a great way to get teens to participate in a qualitative study, particularly when you're doing interviews, for them to take pictures and then describe the picture. That was a really thoughtful and successful means in that study to procure those interviews.
Okay, so then there's some... Criteria for judging scientific rigor and qualitative research and there's one quote at a qualitative research conference recently that I like says rigor equals rigor mortis. So quantitative researchers refer more to rigor, qualitative again it's much more open descriptive narrative.
Don't put a lot of rules on ourselves because that would hasten discovery. So the emphasis is on the resulting narrative. The methods should make sense. The researcher needs to explain decision points in their analysis and in their design of their study. It's not a straightforward recipe and it's okay to make changes.
The study I did on in vitro fertilization, I collected stories off the internet, which is a different, and that was okay. I made a justification for why I did that. and the resulting narratives, again, the data was successful. We look at credibility, the truth of the findings as judged by participants or others within the discipline, so you can show your results to some participants and they can comment on whether they think that has truth to it.
Auditability is the accountability as judged by the adequacy of information leading the reader from the research question and raw data through various steps of analysis. to the interpretation of binding. So auditability is really has the researcher provided an in-depth description of the themes, has the research provided quotes to support that, has the researcher provided enough decision points, enough description so that the reader can audit or the reader can come to some agreement as far as what the the researcher has done.
Fittingness is faithfulness to everyday reality of the participants. Described in enough detail so that others in the discipline can evaluate the importance of their own practice, research, and theory development. And then confirmability as the findings that reflect implementation of the above, credibility, auditability, and fittingness standards.
The other thing we need to focus on, this is the last thing, is ethical considerations in qualitative research. So this is done in a naturalistic setting. There's an emergent nature of the design so as you're doing an interview things come up that maybe you didn't expect.
In one of my earlier interviews, when I was, one of my earlier studies, I was asking about dating and this idea of sex parties came up which I did not expect and it became as actually a side study. So you kind of have to try to prepare for, when I'm talking about dating, I had prepared for conversations about or people becoming upset. So I built into the design of my study protection of my subjects such that I had a counseling agency available during all the times when I was doing interviews in case that was needed and I also had referral information for my participants.
The researcher-participant interaction is ethical consideration to prepare for and then the researcher is really an instrument because you're doing the analysis. You're looking at the narrative, making summary statements, finding the themes. So you need to make sure the researcher is not biased. I can give an example in that online study where I collected narratives, existing narratives in open online databases.
So that was a real ethical consideration and a few of the editors sort of called me on that. I did go back to the nurse ethicist I worked with at Boston College and she said that by using that data I was not posing any additional harm to participants more than what they had already posed by posting it publicly on their own. In fact, I de-identified it so I actually lessened that risk.
And it was essentially I only used... data that was posted on an open online forum. I didn't go to any closed online forums to collect those those stories. So in essence it was not a human subjects research in the end when I actually went through the IRB it was exempt because it was no different than say reading a newspaper or any other public document.
So that's a whole other interesting discussion about the ethics of online data collection and how we go about that. Happy to discuss if any of you are interested. And this was a long one, and that does conclude Unit 2.