You need to make changes to the collected data for yourresearch‐ For example, to create new ways to group your respondents forcomparative analysis•There are primarilytwotypes of re‐coding that you wouldencounter in this unit‐ Creating new variables that allow you toclassify/group your responsesdifferently‐ Usea different set of values to code an existing variable(often to createdummy variables)44Data manipulation ‐ re‐codingRe‐coding example: create new variable “age group”•The new variable contains two categories•Young (<=35) is assigned the value 1•Middle aged and older (> 35) is assigned the value 245SubjectAgeIncome129$40,000245$36,000422$16,000541$98,000633$60,000722$24,000933$55,0001045$80,000SubjectAgeIncomeAgeGroup129$40,0001245$36,0002422$16,0001541$98,0002633$60,0001722$24,0001933$55,00011045$80,0002
5/09/20181646Re‐coding example: re‐assigning value to createdummy variables•current coding‐ Domestic – 1‐ International ‐ 2Nationality Variable•Re‐coding into a dummy variable‐ Domestic – 0‐ International ‐ 1Subject NationalityIncome11$40,00022$36,00042$16,00052$98,00061$60,00072$24,00091$55,000101$80,000SubjectNationality IncomeNewNationality11$40,000022$36,000142$16,000152$98,000161$60,000072$24,000191$55,0000101$80,0000Quick recap about descriptive data characteristics47•Categorical variables: counting and tallying–data distribution (frequency and bar chart)•Metric variables: the “average” behavior–Distribution (histogram)–Mean vs median – Outliers•Data manipulation - re-codingToday’s agenda48•Data preparation–Coding (including the concept of dummy variable)– Cleaning•Getting to know your data – descriptive data characteristics–Categorical variables: counting and tallying (data distribution(frequencies and bar charts))–Metric variables: the “average” behavior (data distribution (histogram), mean, medianand outliers)–Data manipulation - re-coding variables•Intro to hypothesis testing
5/09/201817Recall that most research questions are about variables and their relationships9/5/201849•Relational– RQs that are about the relationship between two different constructs/variables•Comparative– RQs that compare two different constructs or measures of the same construct across different groups or scenarios (often between 2 groups)Now you have the data, how do you “answer” a researchquestion (RQ) like the one below?9/5/201850•RQ: do male and female studentson averagediffer in theirgrade expectation for MKF2121?•BTW, what are the variables involved?We already have the data for the variables collected throughthe mini survey on Moodle9/5/201851•Gender – Q9•Grade expectation – Q7
5/09/201818How to answer the RQ? – compare the average gradeexpectation of the two groups9/5/201852Conclusion – male and female students on average differ intheir grade expectation for MKF2121Mission accomplished?