Impact of Body Shaming on Depression Literature Review

Dr. Alex Casteel – Research Methodology & Design

Click on this website then click the course tab, after EDCO 745. module 2 you will see quiz, which is not a quiz but a topic,

Quiz: Pick Topic

Data Screening Assignment

Literature Review: Outline Assignment

Installing PROCESS in SPSS

Reading info for course—

Qualtrics XM: The Leading Experience Management Software

Reading info—for topic I like

Depression, as measured by DASS-Depression (a continuous variable with interval level of measurement)

EDCO 745
LITERATURE REVIEW: OUTLINE ASSIGNMENT INSTRUCTIONS
OVERVIEW
• The goal of this assignment is to get you moving as quickly as possible on your reading
of research articles. It would be best if you read peer-reviewed research articles at a pretty
furious pace (two a day) to write with sufficient knowledge.
• The literature review is a review of the literature. It is where you provide the
information you have gathered on your selected themes and begin constructing the case
for your specifically designed model.
As you read, consider:
• What is the existing literature out there on my selected themes/variables?
(i.e., 1. You select depression as one of your variables. AND 2. Search existing
literature/studies on depression (begin by gaining a broad understanding of depression,
then begin looking at what particular studies have been conducted on depression,
independently and in relation to your other selected variables)
* Study each research article you read:
1. Take note of the literature they have examined to design their study, but DO NOT
CITE FACTS FROM THEIR LITERATURE REVIEW UNDER THEIR NAMES; go to
the ORIGINAL SOURCE
2. Look at the methodology behind their study, examine their specific procedures and
what tests they conduct
3. Take note of how they report their results
4. When you are reading through their discussion, note the way they discuss (perhaps
their rationale or what sources they refer to)
5. Look at their limitations and future research
• Throughout you are reading, ask yourself where there might be “gaps” in the literature (things
the existing literature has neglected to examine/research) – this will inform your study.
INSTRUCTIONS
Construct an outline for your Literature Review:
• This outline must be thorough and include information from the articles you
have read; you must cite each of them, in APA format (refer to APA
handbook and guidelines)
• The outline of your Literature Review must have at least 5 major headings
• This must include major themes as well as integration between themes
Page 1 of 2
EDCO 745
• Note: You may resubmit this assignment a second time after you receive
feedback
General format:
I. Introduction to entire literature review (must be informative and concise)
II. Theme 1 – This is ONE heading (you need at least five)
A. Detailed info on Theme 1 (include information on reviewed articles, include APA
formatted citations – see APA Handbook and Guidelines on citations)
B. Detailed info on Theme 1(include information on reviewed articles)
1. Even more detailed info on Theme 1 (include more detailed information on the
specific article mentioned above)
2. Even more detailed info on Theme 1(include more detailed information on the
specific article mentioned above)
III. Theme 2
A. Detailed info on Theme 2
1. Even more detailed info on Theme 2
2. Even more detailed info on Theme 2
IV. Theme 3 (continue with more detailed information on Theme 3 as outlined above)
V. Theme 4 (continue with more detailed information on Theme 4 as outlined above)
VI. Theme 5 (continue with more detailed information on Theme 5 as outlined above)
VIII. Integrate and Synthesize the above Themes
V. Conclusion of literature review
Note: Your assignment will be checked for originality via the Turnitin plagiarism tool.
Page 2 of 2
EDCO 745
DATA SCREENING ASSIGNMENT INSTRUCTIONS
OVERVIEW
Data screening is fundamental before any analysis. The basics of data screening are outlined in
your Warner text. Data screening is used for two major purposes: removing bad cases (missing
responses or nonresponse patterns) and testing the assumptions (distribution of quantitative
variables, sample size requirements for analysis, removal of small groups) of the intended
analysis. This is one of three data screening assignments you will have in this class. This
assignment is a general data screening exercise intended to prepare you for the remaining of the
rest of the term.
INSTRUCTIONS
• AFTER OPENING THE SPSS FILE (CLICK ON IT, IT WILL OPEN)
• NOTE: THERE ARE VARIABLE VIEW AND DATA VIEW
1. Choose:
• 2 categorical variables (often represent naturally occurring groups or categories
– i.e. gender)
AND
• 2 quantitative variables (provide information about the magnitude of differences
between participants, in terms of the amount of some characteristics – i.e.
anxiety)
*For further questions regarding categorical versus quantitative variables,
see Applied Statistics I, pp. 31-33
Page 1 of 2
EDCO 745
2. Go through the data screening process outlined in Applied Statistics I & II.
•
•
•
•
•
Frequency Tables – Applied Statistics I, pp. 37-50
Histograms- Applied Statistics I, pp. 103-115
Boxplots- Applied Statistics I, pp. 115-120
Bar graphs – Applied Statistics I, pp. 100-103
Scatterplots- Applied Statistics I, pp. 126-127
Write a data screening story consistent with Warner’s instructions (- Applied
Statistics I, pp. 37-50).
• Make sure you include both your output:
It should have:
1. Frequency Tables for all four variables
2. Histograms (2): One for each quantitative variable.
3. Bar Graph (2): One for categorical variable
4. Boxplots (4): Two Graphs for each Category (e.g., gender X
anxiety and gender x depression)
5. Scatterplot (1): Quant x Quant
Note: Your assignment will be checked for originality via the Turnitin plagiarism tool.
Page 2 of 2
Your topics should be based on something you’re interested in. But it is limited by the data. Obviously, if you have interest and that interest will then carry through and will help your passion and trying to understand it. Also learn anything about any of these things. Depression, personality, shame, anger, loneliness, and religiosity will be useful in developmental for you going forward in whatever path you’re taking in the, in the ministry or helping field. All of these factors can effect or it can be a part of what you’re doing in the future is limited by the data. You can’t have a topic that isn’t in the data that we have. So I recommend that you limited on these things. If you are getting articles about other things, those other things are not going to be useful for you going forward. It will not be helpful for you to be limited by those other things. So or to be reading about other things. So stay focused on the data, pick the topic with, then this is a topic you’re going to spend all the whole term on writing about analyzing and doing data screen. Do the screening is foundational. It’s what we start with from the beginning when we get data in and we first want to make sure that our data is good and that we do it for two purposes. First is kinda make sure we have decent data that something’s not wrong. For example, if we’re looking at, if we’re looking at relationship status and we’re only looking at males who are single. We want to make sure that our sample size is large enough that we had enough people look at males single. So we’ll look at relationship status. And we’ll look at gender, will say, oh, well we only had 20 males while we can’t, we’re limited in the quantitative analysis, we can do single males if we only had 20. So that’s one thing we have to have enough data to do what we’re trying to do from analysis. Another thing is you’re working to define or figure out, did people attend, Are there did they answer all the items to the items you need for your study? And you start filtering and determining, Can you do the analysis? Do you have the data in the data have the quality that you want? You look at you need for this exercise. And this is the exercise you’re not doing it pretty specific analysis. This is two gets you experience with the data. You want to categorical examples would be gender, relationship status. You might use, you might use denomination or faith. You, you can pick any number of things. Demographic sat, and two quantitative measures. Those quantitative measures are based on my depression and religion are depression and faith. And that faith might not faith in the categorical but religious commitment. And that’s the religious commitment, inventory or awareness of God, a subscale Du Sai, or depression and shame. So you can pick any kind of variables that are of interest to you. So what you’ll see in a data screen and I’m going to do a demonstration on the next slide. But what you’ll see is we put. Kind of page numbers where you can go and find how to do these things. But I’m gonna go ahead and demonstrate each of these on the next slide. This is the, this is the dataset, this in, in this course. And what you’ll find is we, we have in SPSS, we have a variable view and we have a data view. This date abuse where the actual data is this variable view or the list of the variables. And you can see those variables will map from, if we see our demographics I demonstrated graphic start down row one and they go down 2j, belief in God right here, rho 12. And then we start and actually we have religious daily life. What religious daily to what degree someone has their religious, and then we have whether they’re sexually active or not. And then here we go to we have anger and then we’re variable star in your field start at 15 or your scale started 15, and they go to 46. And then below that, what you’ll find is you’ll find a number of the individual items for each of the scales. If you want to know what a skills about, you’ll be able to do individual items. You’ll also notice down here there’s social media. Do they have they use social media? And then social media total hours on average. So that might be interesting to you as well. So this is, this is our data. Now, if I’m gonna do data Screening, What I’m gonna do is I’m going to take a look at two categorical M2 to categorical and two quantitative variables. So I’m doing, what I’m gonna do is I’m going to first do data analyze frequencies. And I’m going to bring over, I’m going to bring over, let’s see, race. Let’s, let’s see, I’m going to bring over race as you can see. And so that’s, it will show you a little trick with this. If we right-click, We can go variable names and I go race, age, that kind of thing. I’m gonna do race. And then I’m going to do relationship status. And I’m going to look at age. And this is a church attendance MSE. How often people attended church in the last year? So those are the measures I’m going to bring over and I’m going to do, if we’d look at this, move, this so you can see it a little better. What you see is statistics, I’m honored and statistics at this point, what I’m mostly interested in is frequency tables. And then I’m going to go charts I want. And C, This brings up this. I’m gonna get a histogram and I’m looking for histograms mostly in my quantitative variables. That’s first and we click, okay. So what that does is that opens as output window and I have an output window in this dataset, there’s 1300 people. And in this dataset, we’d look for either analysis that I can do. So if I’m doing data screen, the first thing I’m gonna do is I’m going to look at frequencies of my categorical variables. My categorical variables, as you can see, this is a predominantly White sample. We have African American in the sample, and then we move down and we have Asian and some Latino, but then we have some others as well. So what I would probably do is say, am I going to do a comparison between races? Look at these sample sizes. I would say the only races I can compare, our white versus black. Because of the association, we have 5.2.1, almost 5-1, about 4.5 times the one white versus black. That’s okay. I can’t really compare white to Asia because now I’m at 15 to one and y to Latino. I met twenty two, twenty three, twenty four to one. So you see, I have to think about that ratio. What’s the perfect ratio? Great question. There’s no perfect ratio. It depends on the type of analysis. If I’m doing this massive analyses with tens of thousands of data, I can, I can say it’s 5%. If I’m doing an analysis where I only have, only has 300 cases, I would say it’s 20%. Probably is where I need to be. Here. I’m, i go into the 15 to 20% and that really puts me The only, if I’m looking at white versus black is the only comparison I can make as far as that goes. However, I could go black versus asian versus Hispanic or Latino. So that’s the way you could kind of look and say, what do I want? So if I said, I’m going to do white versus black than what I’m going to do is I’m only going to select, I’m only going to select white and black. Now next, interrelationship, as we see, this is, this is primarily 70% of the people are married life partner. Alright, so I’m going to then, I’m going to look at just married people. And so I’m gonna end that has 70%. So I’m, I’m looking at two things. I’m looking at white versus black. And then I’m looking at married people only. So notice I’m not going to do a comparison, but I’m going to focus only on married life partner and people. That’s my focus. And I’m going to compare race variables. And then we have age and see, age has this frequency table. So that’s nice. It’s long and it goes, look at we have someone 93. I said an interesting thing. That’s something I think I want to get rid of it so far out of range that I think I want to get rid of that 93 and keep just up through 76. Up through 76. My guesses someone put in the year they were born. Let’s see, what else do we have here? What is your age? And 93, someone put a year. They’re born. Now let’s look at church attendance. Church attendance. Look at this. We have samples and then we have people work at here. See these, check this out. I have I have some data here that I suspect is problematic. These looked to me like years. That’s what they look like to me. And then 777 looks like a nonsense entry. 600s possible. I don’t know what that cases but it looks to me like I certainly have some outliers and I have some nonsense and entries in, in Number, how often they attend church? So I’m going to get rid of those nonsense entries is what I’m going to do. So I’m going to, I’m going to just stick with 365 people with ten. So I’m going to decide to get rid of these what I would describe as extreme cases. And that’s from 600 to 2019. Reason we get rid of those extreme cases is is not just about the distribution, but it’s practical. Can someone attended church to thousand times in a year? Well, I think that’s the last time they were injured. But the answer is no. That would be six times a day. That’s possible. It’s possible for there about certainly possible if they were doing prayer sick time six times a day, that’s possible. So I think though, I’m going to remove everyone who who has more than one time a day. Now that’s still going to create some outliers worry those people at 365, but I think that’s going to be so on church. I’m going to remove and notice I’m making an assessment that has happened is that logic is not on the distribution was all. I’m not really even assessing the distribution. So church has gotta be less than, and in this case I would say less than 400. So if it’s less than 400, I will choose to keep the church folk so far. Relationship status, race, I’m just getting white and black. Relationship status. I’m going to, let’s see. And let’s say I only picked one variable, zed, right? I got age. How often they attend church, although as to relationship status in there. So a ten and age so I got age is going to be 76 and below. So I’m getting rid of one on age and I’m gonna get everybody who is above 365, some get rid of those. And I’m only going to include white and black and my sample. So how do I do that? So this is the first phase. This is, I’m going to get rid of also look at this. I have someone who says they’re six years old. I have to get that person sexual assault, someone did not say did not affirmed that they were 80. So that’s six year old person has to go to. All right. So that’s what we’re gonna do. I’m going to go this route. And I’m going to say, all right, data select cases. And this is my select if condition. And then I’m going to go first age. Let’s see, I’ll show you how this works. Let me get this over here. This is another dialog box, as you can see. And so this is a, I’m going to go, age is greater than 17. And Age is less than 90. He’s got it. All right. So that’s that’s one. And so I do this. And so I’m going to do this first and go continue and say, okay, what does that look like? I’m going to do this now watch what I do frequencies again, allies, descriptives frequencies. I’m going to look at age. And now I have my age and I don’t have my six-year-old and I don’t have my 93 year old. Excellent, excellent. So now’s next. I’m going to do data select cases. Again. And this time I’m gonna go and Let’s see, church attendance is less than 366. And I got click Continue. I click OK and analyze descriptives, frequencies. Church attendance. There we go. And what do we have now ours our church attendance as age still looking good. 249 to 365. Beautiful. Beautiful. We’re doing good now. We have age. Now let’s just look at married data select cases. If, and let’s say this is rho status. But let’s say I don’t know what number marriages. So I go like this and handle variable information. And I go in here and I find one is single, five is married life partner. Excellent. So now I know what criteria to use and you go like this, equals five. And now we have that. And so we do continue. We click okay, and we’re gonna do frequencies again to check it. Descriptives frequencies. Alright, and there we go. And now we go and hear our races getting less than here. And what we see is a relationship status. Let me see. This is a quick way to We only have one relationship status. And here it is 861 valid like partner. That’s it. There’s only one by perfect. And now let’s do race. Data, select cases. If. And then we go to, and I went go race. Much loot, display variable names. There is race. And we go, What does race? What, what is race? And so, okay, let’s look at the information, make sure we have it right. White and African-American, I could do this. That race equals one or way race equals two. White has one mascot. Race, or race equals two. Or I could do this. I could do races less than race as less than three. We click continue and we click OK. And now we do analyze descriptives. We do frequencies. And we go OK, and we go, wow, 784 white African-American. Look at their 21 to 76 married. And we only have white and African-American. So that’s pretty good. This is our first wave, our first wave of data screening, how data screen is. I’m going to move to the next thing. I’m going to move to the next thing. So we have gone we have gone from we now only have seven hundred and eighty five, eighty four people. And if we scan all the way up here in our original frequencies, what we see is we see that we did have 131306. So after we’ve done our initial data screening, we have this another thing that the thing about a set of services that we need all we didn’t do our services stuff. Yep. Oh, no, that’s there. It’s that down there now we’ve got our sort out up. How often do you identify that’s gender, age looking good. There’s age, that’s a nice distribution. Relationship status, what we’d expect, how often do you ten services. Now as we see, we’re still very skewed and are attending religious services. So we have to ask them questions about what’s meaningful from a religious service attendants. How do we want to categorize religious service attendants? Do we want to categorize people who don’t attend church at all as one category. And so we need to think about because it’s not normally distributed. So let’s go ahead and look at our frequencies to see how we might organize these. Well, I would argue that we want to say the 0 attenders or their own category. So I would, I would read code, recode church attendance into 0 and then I would go, well, what’s my other argument? Well, how about my once a month, folks? Once a month or less than once a month. Wow. What do we have? We have 45 right here, 45 and below. So to me, there’s the once or twice a year, once a month, once a quarter. So we had to decide, I’m going to go that people who attend the zeros, and then I’m gonna go people who are 12 to one to 12. Budget aside, what’s meaningful as far as that goes, that’s what you have to decide. And so I wanted to go and that I’m going to go I’m going to go more than once a month. And then I’m going to go to right about here, 50 times a year or more. And so that’s what I’m gonna do given what we have. And I think 50 times a year more is fine. That’s once a nearly once a week or more. We have a nice big measure there. So what I’m gonna do is I’m going to transform, I’m going to recode into different variables. And so I’m going to create a new variable. And so I’m going to go to church attendance. And there we go. And I’m going to have a meaningful category in this case. And what we have is this is church recode. And this church attendance. And then I’m going to change and that gives me an old and new buyers. So I’m gonna go 0 is 0 and add that. Then I’m going to go one to 12. Ad is 11 to 120. I’m sorry, one. I’m going to do a range of one to 12 is one. I’m going to change that. So one to 12 was one. So that’s once once a year to once a month. And then what I’m gonna do is I’m going to go range 132, 3613361336. So that’s three times a month. Three times a month in that case. And that’s going to be two and add one to three times a month. And then we’re going to go to 37 to 300. And that’s going to be 3x. And that’s going to be at and that’s going to be more than three times a month. That’s what that’s going to be. Alright and continue and click OK. That looks good. Old and new values. And then I do church recode. And now guess what? Descriptive frequencies. I’ll get rid of this one now. And church ricotta is at the bottom. It just added a new data. Might go like this. And now we look at our church attendance. Now notice I don’t have good value labels here and I got a negative one. I wonder all cuz it’s a histogram, not a bar chart. So we would rather see a bar chart in this case. Because it’s now Categorical. It’s, it’s ordinal. But as categorical Now, look at this. Most of our people are attending 12111 once a month to three times. Once a month to what did we say here? What was that? Let me look at her. And if I wanted to see what I said, I just got my log and I go 13 through 36. So that’s once a month to three times a week. So the question is, should we recode this differently? And it looks like we should, doesn’t it? So if we go 13 to 13 to 24. So let’s do transform. I like that re-code into saying in a different variable. I checked this, I look old and new values. I don’t like the distribution there. So I’m gonna go 13 to 24. And I’m going to change that. And that’s going to be my two. And then I’m going to go, this one is going to be 25, 25 to 36. And then I’m going to change that one. So it’s going to be three. And then the last one is going to be 37 through three through 400. And that’s going to be a four. Add. I like the way that look some of you who continue new church attendance, click again, I’m changing that must do that frequency and let’s do, let’s just do that. Let’s just do our new church, recode them and get rid of all others. Then let’s do charts. Let’s do a bar chart. Because now it’s categorical and categorical bar charts are better. And let’s see. Wow, we see what happened. I switched it when I did the wrong rate, S4 looks good. Three, isn’t that, that meaningful to its one, that the category is really, really beefed up here. And so that’s the question, how do we split this category? And so I’ll have to go back to church attendance and see what we have here in our and our church attendance. Let’s see. 20 is a big number, 12. So if we go if look at that tennis 56, so to me, if we go 121211, that gets a very big number. That’s still a huge number. And then we went and we went 12 2312 once or twice a month, 12 to 24. But that’s what’s going to be a big number. That’s a lot of people. If we see once, twice, three times, I want a tendency to, to keep that number where it is. And let’s look at our code and see if this makes sense for us. For 357346141. What I’m doing now is, you know what? I think I’m going to combine 234, not attenders, infrequent attenders, frequent attenders. That’s what I think I’m going to do. So I’m going to do transform. Recoat into different variable. There we go again. And old and new variables. And I think more than once it, more than once a month, we’re going to say is frequent. If we say more than or do we just combine one through 12? So this is our big category right here. And so the question is, how do we combine these? And you don’t want, it strikes me that maybe we combined 23. So we go 13 to 36, change and we get read removed. And we add that to three, and we changed that. And we continue. Let’s take a look at our distribution again just to see if now we’re in a better comparison. I’m trying to get these folks, you know, how I describe these folks in a way that’s meaningful and there we go. And NIH are act. So now I have this big group who tends once a month or less. And then I have the next two groups are sort of like what we’re doing in our knowledge. I like what we’re doing here. So I would say we’re doing pretty good. Alright, so now I have church tent. So now what I’ve done is I’ve taken church sentence and combine in and taken it to a new variable. And one of the things I’m going to have to do though, is I’m going to search by window and I’m going to go, okay, where do I want to go? I want to go to my data set and I want to go down to church attendance. I would describe these variables. And so I go to values. And so now I say, well, 0 is, do not attend church. Do not attend church. That’s 0. Add. Why? Attend? Once once a month or less. Okay. Once a month or less, Add to. And I forget what they were and I may go to my output. Two is ten once a month or less. What is 22? To forget the off time and let’s just say it’s more than once a month. But less than. So lets go. This is this says 13 to 13 to 2424 and add and then three I think is still okay. Let me figure out what that was and I forgot what I had done. And so when I go back here, I’m going to output 123. Let’s see the recode 11213 through 3637. Ok, 13 to 36 RI. Let’s go back to my dataset window. Go back to, to the data and go here values once a month or last 133613 to 30 secs times a year and change. And then three was three was more than three times a month. And that’s three and we add okay. I’m gonna clean this up a little bit. I’m undoing all lowercase. I think in this case, I think It’s a sentence structures on change. And there we go. We do not attend church. Alright, to do change. That’s good. Looking good here. Alright. This is excellent. And now we have church and we change that into Rico. And that’s an example of a recode. This. And so that’s the first part of data screen. So we’ve done Frequencies and we’ve done some recoding based on those frequencies. Now what I’m gonna do is I’m going to do, and we’ve done some histograms of Dunbar graphs, scatter, and I’m going to do box plots. Box plots is simple. What we do is we do go in here Legacy Dialogs, do a box plot. I’m going to do simple document summaries of groups of carriers, cases define. And here’s my boxplot. And what I’m gonna do in this case is I have a new I’m going to look yeah. I’m not going to use relationship status. I’m going to because I only have one of those, I’m going to look at race white versus black. I mean, that’s going to be my category. And then my quantitative variable in this case, what we’re age and church attendance. I’m gonna bring age over here and see if there’s an age age issue. And so we go like this. And we look in our box plot. And what we find is we have some potential outliers particulate and our Y sample, we have a lot of older white people. And let’s say, I look at this and I say, hey, what? I want to look at this case and see if this case is interesting is there’s something going on. I select something, I select like nano nine, I can right-click and say go to case. And it takes me right to that person in the dataset. And I can go ahead and look and see what’s going on with that person. So there’s something about that person that’s, that’s interesting. And so this is a way that you produce docs boxplots. Personally. Me, personally, I do not do not get rid of people who, who just violate the box plots. I look at other things. Now, scatter graphs are interesting because we have one very skewed, our Anna and our age is pretty skewed as well. So one of the things I’m gonna have you do is this. I’m gonna do a scatter graph. I’m personality. And so scatterplot and so graph and show what a scatter plot looks like and scatter plot. And then I’m gonna do simple scatter define. And this is what I’m gonna do. I’m going to go and say, I want to look at loneliness is my axe. And depression. Because my y loneliest predicts depression. I do a scatter plot. I look and there’s my scatter plot. And that’s kind of how you produce a scatterplot and what you’re looking at here. And this is one of the things you want to look for is the relationship between the, these linear. And here’s the way to figure that out. Do we have a, a moderately or linear relation? I do this, I put a fit line and I say linear. And so I go like this and then I do Linear and I apply. And so this says that I have a relationship of 0.643 r square into account. So loneliness accounts for 64% of the variance in this. So now if I want to see if this is primarily a lender year, I’m gonna do this number. I’m going to go back to here, do a new fit line. And I’m going to do quadratic, and that’s this one right here. So I’m going to do this and I’m going to do quadratic, not lows but quadratic and I apply. What I’m looking for is, does my non-linear relationship quadratic, which is curvilinear, account for a lot more variance than my linear. And that’s one of the things I look for. And I go, oh, you know what? It does not, it’s 0.6 for 4.643. You know what, that’s good. Now, what’s, what’s more? I would say 5, 10% of the variance more. If that’s what you have, then you say, whoa, I might have some problems. I might have to look at some outliers. If I were to look at some outliers, I would look maybe at this case right here. What’s going on here? That this is, this relationship has much lower. These cases and this case here kind of hanging out here. Those would be some cases I would start looking at. Maybe this case right here would be cases I might start taking a look at. And again, you can go to case like this and that takes you to their sees it and to see if there’s some sort of issue of some sort. So all right. That’s it for this video. I know it took a while, but it’s data screen and I know that it’s important for you to get this right. Thank you.