Building upon Residency Project 3, this project assignment first requires you to consider the methodology you might use in collecting data for your study. Specifically, you must address the operationa
To modify and/or improve your understanding here, remember the quantitative technique chosen to test the hypothesis must align with the type of variable that you are using in the study. Wrong techniques can be chosen and will provide a seemingly correct answer; however, the answer will likely be wrong. Using the wrong technique would result in the wrong conclusion for your study results. The steps for quantitative research include (a) state the research problem, (b) define the purpose of the study, (c) review related literature, (d) formulate hypotheses and variables, (e) select the research design, (f) select the population and sample, (g) collect the data, (h) analyze the data, and (i) establish the findings and conclusions. Carefully review the tables to ensure you are choosing the correct technique to test your hypotheses.
After data collection, a quantitative researcher is ready to perform the process of analysis and interpretation. The reminder is the data collection technique does not determine if the researcher is performing quantitative or qualitative research. The problem statement and the technique of analysis and interpretation required to answer the study’s research questions determine the methodology. The quantitative researcher uses specific analysis and interpretation concepts and tools in a typical order. The following sequence should be considered.
Prepare data for analysis. Data must be assembled, transformed into numeric scores, and cleaned, if required. Are the data complete and accurate? Are the data organized for analysis?
Summarize the data and describe your variables. Raw data input is too bulky and messy to work with. Instead, summarize your data into easy-to-work-with compilations. For the quantitative researcher, summarization techniques include frequency tables, figures, means, proportions, and descriptive cross tabulations.
Distinguish if your data are categorical or continuous. A categorical score is a value of a variable assigned to a small number of groups or categories. For example, groups of male students and female student each with high ability or low ability. A continuous score is the value of a variable along a continuum of scores from low to high scores. Continuous scores are called an interval, rating, or scaled score. For example, age from 25 to 65 years old, height from five to six feet tall, or the extent to which an individual agrees or disagrees with an idea.
Prepare for statistical analysis. Become aware of your data’s dispersion of values around the central tendency. Use measures of dispersion, normal distribution, and sampling variation to ensure awareness of range, standard deviation, and variance. Analyze what your decisions must be to support your problem statement and research questions. What will your product be when your study is complete?
Determine the type of statistical analysis. Are you measuring group differences (comparison) or relating variables (correlation)? For measuring differences between groups, analyze unpaired and paired observations and consider the use of the t-test, chi-square test, or paired t-test. For measuring correlations between numeric variables, implement measures of association by considering a scatter diagram, regression line, or correlation coefficient. Use the Table 1 to assist you with your choice(s).
The inferential statistic examples shown in Figure 1 consider continuous and categorical variables while using parametric and non-parametric assumptions. Parametric assumptions include (a) the observations are drawn from normally distributed populations, (b) the observations are independent, and (c) the populations have the same variances. Non-parametric assumption do not include normality, meaning the data do not look like the bell-shaped curve, but include (a) the observations are independent and (b) the variables have continuity.
Table 1
Statistics Tests and Criteria for Choosing the Statistic for Hypothesis Testing
Statistical test/test statistic | Type of hypothesis question | Number of independent variables | Number of dependent variables | Number of covariates | Continuous or categorical IV | Continuous or categorical DV | Type of distribution of score |
T-test (Independent samples) | Group comparison | Categorical | Continuous | Normal distribution | |||
Analysis of variance | Group comparison | 1 or more | Categorical | Continuous | Normal distribution | ||
Analysis of covariance | Group comparison | 1 or more | Categorical | Continuous | Normal distribution | ||
Multiple analysis of variance | Group comparison | 1 or more | 2 or more | Categorical | Continuous | Normal distribution | |
Mann-Whitney U test | Group comparison | Categorical | Continuous | Non-normal distribution | |||
Kruskall-Wallis test | Group comparison | 1 or more | 1 or more | Categorical | Continuous | Non-normal distribution | |
Friedman’s chi-square test | Group comparison | 2 or more | 2 or more | Categorical | Continuous | Non-normal distribution | |
Chi-square | Category within group comparison | Categorical | Categorical | Non-normal distribution | |||
Pearson product moment correlation | Relate variables | Continuous | Continuous | Normal distribution | |||
Multiple regression | Relate variables | 2 or more | Continuous | Continuous | Normal distribution | ||
Spearman rank-order correlation | Relate variables | 1 or more | Categorical | Categorical | Non-normal distribution | ||
Point-biserial correlation | Relate variables | Categorical | Continuous | Non-normal distribution | |||
Phi-coefficient | Relate variables | Categorical | Categorical | Non-normal distribution |
(Creswell, 2002, p. 238)