Quantitative Analysis
Quantitative Analysis at a Glance
Quantitative studies at the doctoral level usually require statistical analysis and hypothesis testing. Very rarely will a study involving descriptive statistics alone be permit to defend. Statistical analysis involves creating a statistical model of any number of variables and determining if there is a difference or relationship between the variables in questions. This decision tree illustrates what statistical test to use based on the type of variable and hypothesis you are trying to test.
​
Hypothesis testing involves assuming that there is no difference or relationship between the study's variables. This called the null hypothesis. Only if the p-value (a measure of statistical significance) is less than 0.05, then the null hypothesis is rejected, thus supporting the alternative hypothesis that there is a statistically significant difference or relationship between the variables.
​
Some of the most commonly used statistical tests are:
​
-
T-test: determines if there is a statistically signifiant difference between two means
-
ANOVA: determines if there is a statistically significant difference between the mreans of 2 or more groups
-
Chi-squared: determines if there is a statistically signficiant difference between the frequencies of categorical/nominal variables
-
Correlation: determines the strength of the association between two or more variables
-
Linear Regression: determines if there is a predictive relationship between a predictor and outcome variable
​
Most statistical analysis for doctoral degrees in the Humanities and some Social Sciences can be handled using the user-friendly IMB SPSS software. More complex multivariate models (such as structural equation models) are optimally handled using the statistical programming language R.​
How will Dissertation Demystified help me?
Our weekly office hours are perfect for discussing how to prepare data for uploading to SPSS. We also offer monthly seminars on conducting statistical analysis using SPSS and even R for studies that require more complex statistics.
​
Our statistical analysis module assists members in understanding and interpreting statistics at a much deeper level. A common pitfall that is endemic in so much research is the over-reliance on p-values without reporting effect size, which is a measure of the size of the difference between two means.
​
We also provide member with the resources to be able to tackle more complex, multivariate statistical analyses to better leverage complex datasets and reveal more nuanced relationships between variables.