DATA ANALYSIS

DM99 Consulting was founded by two geoscientists, and one of our core specialties remains providing  data statistical analysis and results. As we often say at DM99, nothing is too complex, and we are always ready to work with you.

As with qualitative research, studies with quantitative methodologies have a number of possible designs, each of which must be articulated effectively in your research questions (and hypotheses), variables, and testing plan in order to ensure robust results. We can assist you with developing a testing plan and performing your full analysis for each of the below research designs–and we can also help you determine if additional testing is needed to guarantee compelling findings and faster approval. Our experts are proficient with virtually every statistical method and test across a broad range of statistical software packages, including SAS, STATA, R, ArcGIS, Tableau, Power Excel and many others.

• Descriptive: Descriptive analysis is, on its own, not typically considered robust enough for doctoral-level research, because no relationships are being examined or inferred. That being said, it is important in terms of providing a basic summary of your sample and dataset. This is done through measuring, for example, either frequency and percentage (for nominal variables) or mean, median, and standard deviation (for interval variables). Our statisticians can assist you with this initial statistical analysis prior to orient your readers before completing the more rigorous analysis necessary to ensure your results are ready for final review and approval.
• Correlational: While correlational research is also relatively simple, unlike descriptive studies, correlational studies do have both independent and dependent variables. That being said, some of the more critical methodologists and reviewers at the major online universities will often press for a more sophisticated research design and analysis. For researchers seeking statistical consulting help completing their correlational testing, we can perform all necessary analysis using the appropriate correlation (Pearson, Kendall, Spearman, or Point-Biserial).
• Causal-comparative: While the causal-comparative design is similar to a correlational design, it goes beyond simply identifying associations between variables. Researchers who select a causal-comparative design are interested in more directly comparing groups, to determine whether an independent variable affects the dependent variable (or outcome) for these groups in terms of effects, causes, and consequences. While causal-comparative studies cannot fully prove causation, they can point to the need for a more deliberate (rather than ex post facto) analysis. Our statisticians can perform all necessary inferential analysis for your causal-comparative study, including the chi-square test, paired-samples or independent t-tests, and ANOVA or ANCOVA, as appropriate. We can also address any potential issues of internal and external validity that may arise from completing statistical analysis for pre-existing conditions.
• Quasi-experimental: Studies with this type of design involve actually conducting an experiment and analyzing the collected data (rather than working with a pre-existing set of circumstances, as in the above designs). This design remains quasi-experimental, however, because of the lack of random assignment; the groups themselves are predetermined. Because of the presence of an experimental and control group, however, the design and thus the analysis are more robust. Here, too, inferential statistics are appropriate, as well as regression and/or multiple regression analysis.
• Experimental: For truly experimental designs, random assignment is used to determine the experimental and control groups, in order to prevent any other possible factors impacting any differences between the intervention and/or variables being tested. Again, inferential statistics are required to determine the impact of the independent variable or variables on the outcome. Our statistical analysis team has extensive experience with both quasi-experimental and experimental studies, and can complete a full analysis often in as little as 2-3 days.