click below
click below
Normal Size Small Size show me how
Evidence #6
test 2
Question | Answer |
---|---|
quantitative research designs: experimental vs non-experimental | experimental is actually changing/manipulating something. with non, nothing is manipulates, you are just looking to see what the variables are (strictly descriptive) |
non-experimental: types | descriptive, correlation, case-control, cohort studies. |
Descriptive | portray a phenomenon as accurately as possible via statistics instead of language.usually involves a tool/survey/something with numbers |
correlation | describe possible interrelationships among variables. |
case-control | usually retrospective (the care has already been done). compare one group with situation to another without. |
Cohort studies | follow group of subject longitudinally over time to describe or predict. |
quasi-experimental vs experimental | answer questions involving prediction and effects of manipulation. difference is the amount of control. |
Experimental | ALWAYS includes an intervention group, control group, and random assignment to groups. (has to have all 3 of these) |
Quasi-experimental | lacks a control group and/or random assignment. still manipulates independent variable. can have two groups, but not randomly assigned. |
quantitative research design types | pretest-postest; quasi-experimental repeated measures design; survey studies; comparative studies (retrospective vs prospective); methodological studies; secondary analysis. |
experimental pretest-posttest design | measure what we know (r), teach (O1), another test (post-test - x), the clients after (O2). looking to see if the information makes a difference. |
quasi-experimental repeated measures design | how long do participants remmeber information that I teach them? O1, O2, O3 (measure, measure, measure), X (re-teach), O4, O5, O6 (measure, measure, measure). |
comparative studies | retrospective vs perspective. what is my rate before an intervention, apply the intervention, then see how the rate changes. |
methodological studies | measurement design. want to institute a new measurement of something. still looking at the same type of phenomenon, just proposing a different way of measuring it. |
secondary analysis | used especially with national data. utilizing information that has been gather, then reanalyze the data based on a particle finding. |
internal validity | controlling for things that make the sample different. downfall: can't generalize. strong: you have the understand that the data is very consistent for THAT population. means you have a homogenous sample. |
external validity | looking to control for MAJOR variables. looking to increase the validity, b/c you are moving to more of a heterogenous group. (larger sample to washout bias). strong: b/c sample is really large. |
hypothesis: purpose | gives us an understanding for what the researchers think the variables they are measuring are going to be able to tell us. |
simple hypothesis | includes 2 variables only. most commonly give you the null hypothesis. |
complex hypothesis | looking at > 2 variables |
null hypothesis | proposes no difference between the groups. A = B. allows us to look at both ends statistically |
research hypothesis | proposes a relationship between 2 or more variables. |
directional hypothesis | tells us that group A is greater than group B. moves all statistics to one end or the other (based on where we are going with our data). we are sure there is a particular direction this will go. |
non-directional hypothesis | still looking at both ends. uses terms like "great", "less", directional because it predicts that there will be a difference between the two groups and it specifies how the two groups will differ |
purpose of hypothesis testing | rejection of the null hypothesis. |
hypothesis testing | based on rules of negative inference (null hypothesis) - there is no difference in the two group |
data collection | what type of data. how was the data collected. timeframe of data collection. |
types of data collection | physiological/biological measures. observational measures. interviews. questionnaires. records/available data. |
physiological/biological measures | use of specialized equipment to determine the physical and biological status of subjects. tend to be objective, precise, and sensitive. |
observational measures | standardized and systematic plan for the observation and recording of data. must determine what is to be observed and how it will be recorded and coded. |
interviews | data collector questions subjects verbally. open- or close-ended questions. |
questionnaires | paper/pencil instruments designed to gather data from individuals about knowledge, beliefs, and feelings. |
rating scales | list of ordered series of variables. may or may not have underlying continuum. |
Likert scale | measures opinions or attitudes about a concept. a number associated with a level of agreement, frequency, or evaluation. |
Visual analog scale | line 100mm long with verbal anchors at either end to depict opposite feelings. person marks on the line where they are. measure with a rules to "quantify" the data. |
number rating scales | person marks a number where they are. may use verb anchors. most commonly used in nursing (ie. pain scale). |
reliability & validity | ensure the results of the study are "accurate". |
reliability: quantitative | consistency of the measures. different types of reliability (interrater reliability - equivalence; test-restest reliability - stability; internal consistency - homogeneity). |
Cronbach's Alpha | have established itnernal consistency. .7-.9 = very reliable. <.7 = inconsistencies in the way they observed/tested. |
test-retest | .6-.8 = the test is pretty strong. |
validity: quantiative | measure is accurate. comprised of several degrees of several dimensions. types: Criterion-related; construct; content. |
Criterion-related validity | linked it to research or some type of data that supports this being part of the question |
construct validity | have an expert in the filed read it and make sure it is clear. make sure it is at a level that all people can understand. |
content validity | has everything been covered? did we miss anything? everything that is related to the topic is included in the questionnaire. |
qualitative reliability & validity - rigor | credibility, auditability, fittingness, reflexivity. |
credibility | would participants recognize the experience as there own? |
auditability | are measures taken that would allow another person to follow the researcher's thinking. |
fittingness | are the finding applicable in other situations? |
reflexivity | does the author discuss their effect on research process or findings? |
correlation statistics | statistical test to examine how must 2 variables are connected to consistent changes in one another. (r). range from -1 to +1. |
positive correlation | if one values goes up, the other goes up as well, OR, if one value goes down, the other does as well. |
negative correlation | if one value goes up, the other goes down, or vice versa. |
correlation statistics: < .20 | slight, almost negligible |
correlation statistics: .20 - .40 | weak correlation; definite by small relationship |
correlation statistics: .40 - .60 | moderate correlation; substantial relationship |
correlation statistics: .60 - .80 | strong correlation, marked relationship |
correlation statistics: > .80 | very strong correlation; very dependable relationship |