[Music] this week we are going to take a look at um adding on to our discussion about developing our our Concepts and our hypotheses and how we're going to collect data and what types of data are we going to collect by talking about Concepts operationalization and levels of measurement because measurement is often difficult and precise you know it seems like you when we talk about measurement you think well then it's going to be precise because we're measuring something but the problem is that researchers have to try to describe the measurement process so explicitly that someone else could come behind them and do the exact same measurement and with social sciences that's hard like if we want to measure crime in a community well what is it that we're measuring are we are we measuring violent crimes are we measuring drug crimes are we measuring all crimes which would include jaywalking and those sorts of things so thinking about how we measure something uh can be difficult seems easy at the beginning but can be difficult so we want to move away from vague ideas of what we want to study say we we are interested in human trafficking okay that's a great topic there's lots of research that you can do on that topic but it's very vague international human trafficking trafficking in the United States trafficking in St Louis trafficking in keep Gerardo um so we want to narrow it down to being something that we can recognize and measure in the real world otherwise we're going to be unable to communicate how our ideas are really relevant and how those findings can apply to our audience so the first thing we want to think about are are the conceptions and the concepts so we want to clarify these abstract mental images like I I mentioned we want to know um what is the definition of a crime okay think about the conception what is the mental image that we have about something how did we develop our view of that term right so if I say drug dealer the concept that most of us have is an illegal drug dealer right but isn't a pharmacist also a drug dealer so we have to think about what's that mental image just when we say a word in our research what's that mental image that is created and then think about our Concepts right what are the words the phrases the symbols that are used to represent represent these mental images in our communication um some examples that your textbook U brings up is like gender punishments chivalry delinquency poverty those are all things that can come out of our conception of crime so getting that down on paper Okay three classes that scientists measure one is direct observables these are things are qualities that we can observe directly like what color is the average female wearing in a particular class or what are the males in the class wearing jeans or khakis right so it can be something as as simple as that in concept it seems simple but if I say how many people in the class are wearing a blue shirt what is your concept of blue is it would you consider like a teal would you even consider that a blue what about a light blue blue versus a dark blue would you only count like a navy blue color is what you consider blue so sometimes even things that we think oh this will be easy can get difficult when we are trying to either Express that concept or measure that concept then you can have indirect observables and these require um a little more detail so they may be more subtle or complex or indirect observations like if we're looking at reports or Court transcripts or criminal history records there's lots of things that you can pull out of that and it's not something that we have observed ourselves if we're reading a police report we're observing what's go went on in that report but indirectly we weren't actually there and then we can measure constructs and these are theoretical Creations they can't be observed directly or indirectly um they're similar to a concept okay so when we talk about the term conceptualization and I think um with this class I'm going to push pause for just a minute on this chapter a key piece of this class is vocabulary and understanding the vocabulary and understanding how it applies to a particular concept or Theory or premise so this class is more than just learning the vocabulary you've got to know the vocabulary to do well in the class that's for sure but you have to apply the vocabulary the theories the premises so a lot of if you notice on the quizzes in the exams a lot of it is examples because you can just spit back vocabulary to me but the only way that I know you truly understand the theories and the premise and the the concepts behind that vocabulary is for you to be able to apply it to an example with conceptualization so this is a definition thing that's going to apply to um examples it's going to apply to quite a bit that we talk about in the rest of the class we are specifying precisely what we mean when we use particular terms it's of often a set of indicators of what we have in mind and it indicates a presence or an absence of the concept we're studying for example if we are measuring violent crime we would Define that or conceptualize that as offender uses force or threatens to use Force against a victim so we would come up with our conceptualization of violent crime your textbook also talks about indicators and dimensions a dimension is a very specific aspect of a concept crime seriousness can be subdivided into Dimensions um could be um the dimension that we use is victim harm is that physical injury is that economic loss is that psychological consequences um the the specification allows us to lead to a deeper understanding and help us get a grasp of exactly what it is we're studying and as we're going through each of these chapters think about your research proposal think about your topic how will you narrow that down to something that is researchable and how will you conceptualize your hypothesis your definitions um thinking about all of those things as you read through these chapters be sure and be looking at it through the frame of your own research project concepts are abstract and only mental Creations the terms we use to describe them do not have real and concrete meanings think about it when we say what is poverty do you live in poverty what is poverty that's a pretty relative term right for some in the United States poverty is not having an iPhone for others poverty is not having a house and then for others poverty is not being able to feed yourself or your family same with delinquency for some getting sent to the principal's office is delinquency for others they'd be like unless you get arrested I don't consider that delinquency what about strain for some individuals they they turn to a life of crime because they want a particular pair of tennis shoes or they want a TV or they they want something that is of um minimal monetary value and again that's even a relative term right what's minimal for me might not be minimal for someone else other individuals experience much more strain maybe they're homeless but they don't choose to go out and commit a crime so that strain can even be a relative term okay the process of regarding things as real that are not is reification all right so how do we create this conceptualization order okay the first thing we want to do is come up with a conceptual definition what is sces let's say okay we want a working definition specifically assigned to a term that provides Focus to our observations and it gives us specific working definition so that the readers will then understand the concept just a a review sces is social economic status okay so then once we have our conceptual definition what's going to be our definition that our readers will know understand the concept then we have to develop an operational definition how would we measure socioeconomic status so we've got our concept what is what is it okay then we have to decide how will we measure it and and your operational definition spells out precisely how the concept will be measured so the progression of measurement steps is that you come up with your overall concept what you want to study then your conceptual definition and then which is how you will Define it for your reader and then your operational definition of how you will measure that and then you can bring those measurements into the real world you want to make sure you can right so so let's talk a little bit about your oper operationalization say that five times real fast your choices okay this is the process of developing those operational definitions so it moves us closer to a measurement requires us to determine what type of data are we going to collect is it going to be qualitative data is it going to be quantitative data or we going to do a survey would a face to-face interview be better how are we going to measure crime reduction for example so thinking about all those things we have to look at measurement or consider what our scoring is going to be where we're assigning numbers or labels to a unit of analysis in order to represent those conceptual prop properties um we're making observations and we're assigning scores to them um it's difficult and CJ or any type of social science research because basic concepts are not perfectly definable so sometimes we really struggle with this every variable should have two important qualities when you're thinking about you know what are the variables in my research one is they should be exhaustive so you should be able to classify every observation in terms of one of the attributes composing the variable so in other words think of all the options and again sometimes that seems relatively simple what if I ask for gender it's male female right but what if someone considers themselves transgender or what if someone doesn't consider themselves as any specific gender so you you want to make sure that you think about all every possible observation say if I'm asking um I'm sending out a survey to students and I want to know if they're freshman sophomore junior senior that seems pretty exhaustive right but what if they're a returning student or what if they're a grad student or what if they're a high school student taking it for advanced placement so make sure when you're thinking about what variables when you're putting your survey together maybe that's what you want to do for your research proposal think about is it exhaustive have I covered all of the options and then it also has to be mutually exclusive so you must be able to classify every observation in terms of one and only one attribute okay the example your textbook gives is employment status making sure that you can't fall into two different categories now another key component if you've taken um statistics already this is kind of a review is you need to consider your levels of measurement um and there are are four basic levels of measurement the first is nominal um I think of this as labels or names race gender um the state of you where you live your state of residence um they are simply names I can't put them in any order I can't say one is better than the other they are just names assigned names ordinal is a step higher and this is where the attributes can be logically rank ordered like if I ask are you freshman sophomore junior senior if you tell me you're a senior and the person next to you tells me they're a freshman I can automatically infer that the senior has more hours than the Freshman however I cannot tell how many more hours I can just rank them interval gives you meaningful distance between the attributes uh the example your textbook gives is temperature if you tell me it's 52 this morning and then it's 72 this afternoon I know there's a 20° difference okay I can tell and that that's meaningful IQ is another example ratio also has that meaningful distance but it has a true zero point temperature you can have a negative right ratio would be like age you know you can't be negative one right negative zero um the number of priors that you have you can't have a negative number of Prior so these are the levels of measurement and why they're important is because based on this level of measurement determines a number of different things and the the type of data that you can gather um the type of statistical analysis that you can do so the the the higher up right nominal at the bottom then ordinal than interval ratio the higher up you can go the more inferences that you can make right and so there are some implications of choosing a particular level of measurement as I said certain analytical techniques have levels of measurement requirements the other thing is that ratio you can treat that as nominal ordinal or interval but you can't take nominal you can't take gender and make that a ratio level of measurement so you cannot convert a lower level of measurement to a higher one so therefore you want to seek the highest level of measurement possible or that's reasonable um so that it gives you more options the key standards for measurement quality are reliability and validity okay measurements can be made with varying degrees of pre Precision right Common Sense dictates that the more precise the better so the more precise you can get your data the better however it's often impossible to get um complete Precision or exact Precision so we look at reliability and this is whether a particular measurement technique if it's repeated over and over and over to the same object would it yield the same result each time the problem comes up is if even if the same result is retriev D it may be incorrect every time right so reliability does not ensure accuracy and we also have to recognize that the observer's subjectivity might come into play but the thing we wanted to strive for is that our measurement instrument is reliable and that we're getting consistent results each time so there are some methods that you can develop to to deal with reliability issues you can do a test and a retest meth method so you make the same measurement more than once and if you do that you should expect the same response both times you can use what we call interrater reliability where um the you have multiple Raiders and then you compare their measurements and you verify if they're getting basically the same results and then you can do the split half method and this is where you make more than one measure of any concept and you see if each measures the concept differ ly so maybe I'm looking at um students opinion on legalization of marijuana I could do a split half method where I have one question that um presents it in a positive light and another question that presents it in a negative light so you would expect that the respondent would have two different answers um the other is validity and this is the extent to which an empirical measure adequately reflects the meaning of the concept under consideration and we talked about validity in the last chapter are you really measuring what you say you're measuring demonstrating validity is more difficult than demonstrating reliability because how do you know if you're not measuring what you say you're measuring because you've set up your your um research to address the the validity so you think you've got it okay there's some methods with dealing with validity one is just looking at the facei validity on its face does it seem valid does it jive with other common agreements and mental images um Criterion related validity Compares a measure to some external Criterion construct validity is whether your variable relates to another in The Logical expected Direction content validity does the measure cover the range of meanings included in the concept and then just having multiple measures right alternative measures could also um help address any validity issues your textbook also talks about composite measures and this is allows us to combine individual measures to produce more valid and reliable indicators so the reasons for using a composite measure is one the researcher is often unable to develop a single indicator of a complex concept if I'm conducting an interview of someone who's a victim of a crime and I'm wanting to know the impact of being a victim of that crime is there one single question that I could ask probably not okay so we may want to to to use a rather refined ordinal measure of a variable arranging the cases in several ordinal categories from very low to very high okay indexes and scales like scales um may have taken surveys that ask if you strong L agree agree neutral disagree strongly disagree right um that's where you're taking those ordinal measures of a variable and rating asking the the individual to rate them from very high to very low okay you'll also see typologies these are produced by the intersection of two or more variables to create a set of categories or types okay um the example in your textbook is typology of delinquent criminal acts okay time one and time two and so you'll see these a lot typologies for serial killers typologies for um juvenile delinquents uh typologies um it again it's that intersection of two or more variables that create a set of categories index of disorder and your textbook talks about this one this is an example example distinguish between physical presence and social perception okay so the index is created by averaging scores for each measure looking at physical disorder and then looking at social disorder there are some benefits of using indexes um one it can be more valid than just a single question like I said sometimes particularly in social science criminal justice there's just not one question you can ask to really address the the concept so Computing and averaging scores um across all items in the category can create more variation than we could obtain in in a single item a single question two indexes are more parsimonious than nine individual variables so data analysis and interpretation can be more efficient if we are developing those indices so that's everything I have for for chapter 5 the key thing with chapter 5 is really thinking about in relationship to your research project what is my main concept how do I Define that concept how do I develop that operational definition how do I measure that concept and then recognizing there are different levels of measurement so those are really the key pieces that come out of chapter 5 along with thinking about would my research be reliable and would it be valid so that's really what you should focus on for this week and also as you're thinking about your research project