Research Process

Research Process

1. #! Make a Problem Statement

   • most provided as a research aim or objective, is the problem/situation clear?, is the way in which the problem will be addressed clear?, is this research worthy problem/situation
   • a problem statement describes what the situation is the requires changing or understanding and how the study will address this situation or problem

2. #2 Create a Research Question

   • how researchers ask a specific and falsifiable question ? testing it using population/patient, intervention/exposure, comparison, outcome
   • good question includes the important variables of the research problem in ambiguous terms. The answer to the question is only as solid and specific as the question

3. #3 Create a Research Hypothesis

   • the hypothesis is what makes the researcher question testable by specifying a falscifiable relationship between research variables. A prediction of a specific instance of the larger understanding
   • are the research variables clearly stated? , are the research and null hypothesis provided? - "if, then" method, have researchers plan to resist a type 1 error by setting an appropriate alpha (0.05, 0.01?). If not can you assume an appropriate alpha by what test data the researchers are reporting as statistically significant? Are the researchers building disiplinary knowledge by working generating hypotheses based on displinary theory?

4. #4 Operationalize the Measures

   • Reliability
     • the degree of test to consistently measure the same thing if its measured in the same level
   • Validitiy
     • the idea that the scale is measuring actually what it says its measuring, Ex. depression and sadnessor or depression and anxiety
   • R value- closer to one the better, closer to 0 the worst, r > or = to 0.7, indicates strength of tool
   • interrater liability: are all my research assistants getting the same value
     • internal consistency: if you ask questions all questions relate to depression and measure it
       • test-retest: consistency between two tests
   • Content Validitiy: finding words that conform this construct- going to professionals who know alot about depression
     • Criterion Validitiy: measure against gold standard crtieria
       • Construct Validity: happens after years of testing, with mutiple scales of depression and the converge, test with scales of happy converge

5. # 5 Design: General Conditions

   • Internal Validity: did you follow the recipe of the method chosen ? if they skip steps and ignore pieces, less likely to trust results
   • External Validity: are the results applicable outside the research experiment? How closely the population and sample match up
   • EXPERIMENTAL: should use true experimental design RTC to get a high degree of confidence that a cause and effect relationships exsists between independent and dependent varables
   • QUASI EXPERIMENTAL: when RTC is impossible or impractical, the degree of confidence the researchers can have in establishing cause and effect is diminished. Manipulation of indepen. variable is must have
   • NON EXPERIMENTAL: observational in nature, and researchers can never be justified in making statements about cause and effect relationships between variables... only correlations! No manipulation of indepen variable, not random assignment
     • Positive Correlation: one variable increases, so does the other
       • Negative Correlation: one variable gets smaller, other doesn't
         • CORRELATION DOES NOT MEAN CAUSATION: a relation exisiting between pheonomena or things or between mathematical or statistical variables which tend to vary , be associated, or occur together in a way not expected on the basis of chance.

6. # 6 Sampling

   • failure to plan for and to assemble to a representative sample will diminish external validity of the research outcomes
   • failure to do a sample size calculation will diminish internal validity of the experiment by increasing the chances of committing type 2 error
   • Random Sampling: hard to achieve in real life, hard to make sure sampling from whole population
     • Random Allocation: random allocating individuals into group of control and research
   • Inclusion Criteria: were the inclusion criteria appropriate for the question?
     • Exclusion criteria: control piece, person meets all inclusion criteria but may have something esle going on that can effect experiment, variable that can mess with dependent variable

7. # 7 Data Collection: survey, interviews, observation: Hawthorne affect

   • #8 Data Analysis: important that "level" and number of comparisons " match the statistical test used to analyze data

8. ETHICS APPROVAL

   • TEST: Iif the checks of internal validity have been met, we can assume that the experimental resuts have meaning
     • DISSEMINATE AND TRANSLATE: impact factor: a number that says how impactful the journal is, level of impact it has on science, meaningfully coherent? findings clinically significant?