In this project, the requirements are both a statistic analysis and a bit writing as well to fully complete the research question! The research question is: Is there an association between, on the one side, happiness and, on the other side, the level of socialization and the number of meaningful connections? (CYPRUS)

Subject:StatisticsPrice:0 Bought3

Share With

In this project, the requirements are both a statistic analysis and a bit writing as well to fully complete the research question! The research question is: Is there an association between, on the one side, happiness and, on the other side, the level of socialization and the number of meaningful connections? (CYPRUS). The program that have to be used here is IBM SPSS Statistics (no other program!) together with European Social Survey (ESS). Everything regarding this statistics project are in the attached files below. If there is a need for seeing the powerpoints from the lectures, contact me and I can send them! One of the attached files are an example on this statistic project, one of my classmates did it, but did NOT pass the assignment due to the reason there were something wrong with the numbers. I have attached it only because the structure of both the theoretical and statistical together and how it should look. But unfortunately the assignment was failed, so hopefully my friend can also use your service after my recommendation! Best regards, Jennifer
Project Requirements Within the project, you have to answer the research question (RQ) that you received via e-mail. The assignment is individual and you have received your research question individually via university e-mail ending with @student.lnu.se. Please, contact Elizaveta Kopacheva before 25 November 2022 if you have not received the e-mail with your research question. The assignment will not be accepted if you answered the wrong research question (not the one specied in the e-mail). To answer the RQ you will do the following: 1. Dene hypotheses (you can refer to lectures 2–3, i.e., “Introduction to quantitative research method: Concepts and Denitions” and “Using quantitative methods for research”, to do that); 2. Select and pre-process data needed to conduct the analysis (you can refer to lectures 4–5, i.e., “Fundamentals of statistical analysis” and “Basic Statistics”, and workshop 1 to do that); 3. Conduct the analyses to test your hypothesis/hypotheses (you can refer to lectures 4–7, i.e., “Fundamentals of statistical analysis”, “Basic Statistics”, “Simple Linear regression” and “Multiple regression”, and workshops 1-2 to do that). 4. Write a scientic report. Analyses and scientic report in detail Defying hypotheses This part should approximately occupy 0.5–1 page. You can write this part of the report after lectures 2–3 as this is a purely theoretical part of your assignment. We ask you to formulate two null and two alternative hypotheses that answer your research question (the one you have received via e-mail). We want you to write a short introduction into RQ and a paragraph explaining why you expect the relationships formulated in the hypotheses. • Start with Introduction: here, write why the RQ is interesting in relation to contemporary issues. • Dene your hypotheses. Every RQ contains three variables: 1 dependent (outcome) and 2 independent (predictors) variables (DV and IVs). For example, RQ: How political and social trust inuences participation in protests? What is my DV? It is participation in protests. What are my predictors? IV1: political trust; IV2: social trust. My 1 ?? hypothesis will be about the relationship between my DV and IV1. For example, H1.1 (alternative hypothesis): political trust and participation are negatively associated. • Explain what it means and why I expect that. What does a negative association mean? It means that when one variable rises, the other decreases. So, I will specify that my H1 is that people with low political trust participate in protests more. Then, I will explain why I expect that. In short, this is because the agenda of the protests is usually policy-change-oriented. In your assignment, please, explain in the way that the examiner follows the train of your thought. • Specify also the null hypothesis. Please, remember that the null hypothesis is the absence of the relationship and not its opposite direction (H1.0: there is no association between political trust and participation in protests). My 2 ?? hypothesis will be about the relationship between my DV and IV2. For example, people with high social trust participate in protests more (my alternative hypothesis is the following: social trust and participation are positively associated/correlated) as I am more likely to go to a protest if I believe that many people will come. In a nutshell, in this part of the report, you specify the relevance of the topic, your hypotheses (expectations) and your motivation (why do you expect such relationships between the variables). In scientic articles, those questions are covered in 2 sections: “Introduction” and “Theoretical framework”. In those sections, researchers usually provide the scientic context of the RQ and specify expectations. In this assignment, you are not expected to be adept in the eld of studies but you need to explain why it is important to investigate the issue and what your expectations are based on (you can explain it in your own words but in the way that the examiner follows the logic). The report does not have to include scientic references. If you want to include references, provide a list of references. This part of the report tests if you, in part, meet the second and third objectives of the course, in particular, “In academic writing, independently present analyses on the basis of a quantitative method” (this includes correctly formulating null and alternative hypotheses) and “[...] in writing account for arguments [...] in the course literature” (this includes being able to explain the relevance of the topic and to motivate your expectations). Choosing and describing the variables First and foremost, we ask you to subset the dataset, provided for the assignment (ESS9e02.sav) to keep only those observations that are from the country that you have to study in your individual assignment. The country was specied in the e-mail sent to you along with the RQ. For example, if I got Germany: rstly, I will subset ESS9e02.sav by country to keep only observations from Germany. Next, we ask you to choose 4 variables from the provided and subsetted dataset. Refer to the codebook (you can nd the codebook in the folder along with the Project requirements and the dataset) to understand what variables actually mean and how they are coded. Those 4 variables should be the following: 1. DV. In my example, DV is participation in protests, so, I will search in the codebook variable that measures participation in protests. 2. IV1. In my example, it is political trust. In ESS there is no such variable really. There is trust to politicians, trust to political parties. This is not really political trust. So, I will have to rephrase my hypothesis, to specify what I mean by political trust: so it corresponds to my measurement. In your assignment, we intentionally choose concepts that are measured by one of the variables in the dataset. You just have to nd this variable. 3. IV2. In my example, something that measures social trust. 4. Control variable. Please, refer to lectures 4-5 to know what control variables are. Control variables do not change for respondents. It could be age, education, gender. Something, that cannot be inuenced, in laboratory, when conducting an experiment. in comparison, political trust can be changed: we can provide evidence of corruption, so, the political trust of a given respondent decreases. We cannot decrease the age of this respondent. After choosing 4 variables, subset your dataset to keep only those 4 variables. The dimensionality of your dataset will drastically reduce after subsetting by country and keeping only 4 variables (so, you will have only around 2000 rows and 4 columns). In this case, all operations on the dataset will take milliseconds. Next, check how the measurement of the variables is specied in SPSS: often in SPSS, variables, that we understand as numeric (for example, ratio variables measured from 0 to 10) are coded as ordinal. Change this measurement to numeric. In your report, describe 4 chosen variables. For each variable, in particular, specify (1) what this variable measures and if it is your DV, IV or control; (2) what is its scale (categorical (nominal/binomial/ordinary)/continuous (ratio/interval)); (3) min and max values and what they correspond to (for example, min is 0—low trust; max is 10—high trust); (4) range; (5) interquartile range; (6) centre of the distribution (median/mean/node); (7) skewness; and (8) kurtosis. Finish with building a graph of the variable. Here, we do not want to see pure statistics in a table. We want to see that you understand how the variable is measured and what is its distribution. So, it is your independent decision how you will describe your variables. For example, if one of my variables is binary, then providing min and max means that I don’t understand how the variable is measured. Binary variables don’t have min and max. They have levels: for example, male-0, female-1. Female is not max; while male is not min. This does not make sense. The same goes for the range, what is the range of a binary variable? The range does not tell the reader anything about this variable. The same when it comes to the centre of the distribution: mean is meaningless when describing categorical variables. When it comes to graphs, the same logic follows. Does it really make sense to represent a ratio variable by a histogram? A better representation would be to use a boxplot. So, instead of providing pure descriptive statistics, we expect to see in the text your interpretation of these statistics. For example, if the distribution is negatively or positively skewed, what does it mean with respect to your variable and its measure: are there more people in the sample with high political trust or there are more people with low political trust (as an example)? To sum up, in the report, you will have 4 subsections for each one of the variables. The subsections can be titled as, for example, the following: “DV: participation in protests”, ”IV1: political trust”, “IV2: social trust”, and “Control variable: Gender”. In each of the subsections, you will describe your variable, based on their summary statistics: starting with the scale of the variable and nishing with a plot (graph). When pasting tables or graphs, each table or graph should have a caption and each table or graph must be described in the main text (this is a scientic standard). Once again, simply writing that, for example, the interquartile range is 4 and ranges from Q3 (10) to Q1 (6) is not enough. You need to explain what it means with respect to the data: that, in this case, for example, 50% of the sample are people with high political trust (6 and more). In this part of the report, we test if you, in part, meet the following objectives of the course: “critically and independently apply quantitative methods in social science research” (before doing analysis and applying methods, you need to understand your data: otherwise, you will interpret the results incorrectly) and “In academic writing, independently present analyses on the basis of a quantitative method” (including presenting your data as in any scientic article in section “Method and data”). This part of the report will occupy no more than 3 pages including 4 graphs. The description of each variable should be around 0.5 pages. Cite course literature if needed. Refrain from giving denitions and rather focus on explaining your choices. Doing and reporting your analysis Before conducting analyses recode your variables if needed. For example, when we are working with categorical variables, we need to create dummy variables to test the relationship between our dependent and independent variables. Do that if you have such variables: but keep the original variable in the dataset (you will need it for cross-tabs). Don’t bucketize interval or ratio variables: that leads to the loss of information. Explain in the text, how you transformed your variables (if you did) and why. Next, test for a relationship between all your 4 variables (1-to-1 relationship). In particular, do the following: (1) build graphs; (2) apply cross-tabulations if you have categorical variables; (3) apply comparison tests (two-sample t-test) in the case of binomial variables; (4) apply correlation tests. Test for relationships between all of your variables (including relationships IV–Control).1 Building graphs In the lectures, we have presented two main ways to depict relationships between 2 variables: those ways are scatterplots and boxplots—however, in some cases, we can also use histograms and barplots grouped by categories. Choose the best graph type to present the relationships between two variables depending on their types. Each graph must have a caption and must be described in the text. Explain why did you choose this type of graph to represent the relationship. Remember that the type of graph depends on the scale of your variables. For example, producing a scatterplot to depict a relationship between two ratio variables has little (if any) meaning. Correlation or the absence of it can be shown in a graph: use the best type to make your point. Cross-tabs If some of your 4 variables were nominal before recoding them, use cross-tabulations to examine the relationships between the original variable and other variables. Show the results of this analysis in a table and interpret this table in the text. Every table must have a caption and be addressed in the text. Once again, use the original pre-recoded variable to apply cross-tabs. Comparison testing Apply two-sample t-test to compare means of two groups. Report the results in a table, summarise this result in the text and conclude if you found the dierence between two means to be signicant. Correlation analysis Check the correlation between your variables. Report the results in a table and interpret these results in the text. What does a negative correlation mean in regard to your data? For example, people low on political trust are expected to participate in protests more or less? Is this correlation signicant? Once again, we focus on your interpretations. 1Why do you need to do that? There are all kinds of implications in the case when your explanatory variables correlate with each other (including the problem of multicollinearity): for you, however, that can be a nice learning opportunity; as well as an opportunity to detect errors in pre-processing data. Regressions After checking for 1-to-1 relationships, you will proceed with many-to-1 regression: multiple regression analysis. Firstly, include only 3 of your variables: test for the relationship between DV and 2 IVs. Report those results in a table and describe and interpret the results in the text. Is there any signicant association between the DV and one or two IVs? Is this association positive or negative? What does it mean in regard to your data? For example, people low on political trust are expected to participate in protests more or less? Secondly, include all 4 of your variables: test for the relationship between DV and 2 IVs and the control variable. Is the eect of the variables still signicant? How can we interpret the change in signicance or the absence of such change? For example, the eect of political trust became insignicant after adding the control variable into my model. So, participation is explained by age rather than trust. Report the results in a table, describe the results in the text and interpret those results. Based on this, reject or accept H1.0 and H2.0 (null hypotheses). We expect this part of the report to occupy no more than 5-6 pages including tables. Once again, focus on interpretations and relate to your research question and the knowledge that you gain about the population of your country. In this part of the assignment, we, partially, test if you meet all four objectives of the course. This part of the assignment corresponds to the sections “Results” and, in part, “Discussion”/“Conclusions” of any scientic article. The assignment is graded with pass or fail. A report is graded with a pass when both stand: (1) it meets all of the mentioned requirements, and (2) the interpretations are correct.