click below

click below

Normal Size Small Size show me how

# Research

Question | Answer |
---|---|

True experiment | (Randomized controlled trial) compares two groups, two groups being manipulated, with participants randomly assigned to a groups |

Quasi-experiments | (non randomized control trial)- participants NOT randomly assigned, limitation-decrease researcherâ€™s ability to draw cause-effect conclusions, pragmatic/ethical issues can arise, bias selection process can occur, often groups are preexisting |

Preexperimental research | pretest/posttest design, single group is compared before and after, one group, nothing manipulated receive the same intervention, no random assignment Much less powerful than a design with a control or comparison group |

PICO | Population, Intervention, Comparison, Outcome |

Levels I, II, III, IV, V | I Systematic review randomized control trials II randomized control trial III Nonrandomized control trial IV One group pretest and postest (preexpermental) V Case reports and expert opinion |

Randomized controlled trial | 2 groups: experimental and control Participants randomly assigned Intervention applied to experimental group |

Systematic Review | analyzes multiple randomized controlled trials |

Reliability | consistency of measure, dependability of scores |

Validity | assess what it was intended to measure |

Sensitivity | proportion of individuals who are accurately identified as possessing a condition of interest |

Specificity | proportion of individuals who do not have condition |

Prevalence | People who have a particular condition |

Incidence | Risk of developing a condition |

Cross sectional/ longitudinal | CS- single point in time, L: multiple times |

Impact factor | How many times journal has been cited |

Publication bias | Research to have positive outcomes |

Experimental research | Cause and Effect, efficacy/ intervention studies, one intervention, one group control |

Nonexperimental research | cannot determine cause and effect, descriptive relationship, observational and correlation |

Third variable | 2 constructs could be related but the third variable could have caused the relationship (ie cold weath causes flu, or person's bx causes flue (3rd variable)) |

Experimental/ Efficacy all study designs | true, quasi, preexperiment |

Nonexperimental- Descriptive-- all study designs | D-Group-two or more existing groups identify differences, Incid/Preval- occurrence of one or more characteristics calculated |

Nonexperimental Relationship all study designs | R-Correlation-relationship between two or more constructs calculated Predictive-predictors are considered in terms of their impact on an outcome |

Directional hypothesis | researcher has assumption/belief on particular outcome |

Nondirectional hypothesis | exploratory, no prior notion on results |

Quantitative | Focus on confirmation, test hypothesis, objective, deductive ,numbers, is it reliable and valid |

Qualitative | builds on theory, exploratory, subjective, inductive, interviews, trustworthy |

Basic research | questions to better understanding individual concepts ( environmental conditions and stress rxn) |

Applied research | direct to health care ( efficacy of fall prevention programs) |

Type I and Type II error | Type I- hypothesis is accepted even though it is false-occure bc chance Type II- hypothesis is rejected even though it is true- sample size is too small |

When do we accept research hypothesis? | p < 0.05- still a 5% chance that hypothesis is false, 95% chance that it is true |

Categorical variables | autism- girls and boys, gender is a catergory |

Continuous variable | numbers have meaning to one another, higher-- more than something- autism spectrum- age and grade level |

Independent variable | what is being manipulated or compared, intervention (intervention and control--2 variables) |

Dependent variable | intended to measure the result of the manipulation ( interested in communication skills in of participants- communication skills is dependent variable) outcome |

Control variable | remain constant |

Extravenous variable | tracked and then later examined to determine influence |

Descriptive stats | describe study in a meaningful way, patterns that emerge from data |

Inferential stats | techniques that allow us to use study samples and make generalizations that apply to population (descriptive used in the calculation of infer.) |

Frequencies/frequency distribution | -how often something occurs -graph used to depict the count |

N,n | number of participants |

a (alpha) | level of significance |

p | probability value, liklihood of making a type I error |

t and F | t- critical value in a t-test F- critical value ANOVA |

r and r2 | r-correlation coefficient r2-variance in correlational study |

ES | effect size-magnitude or strength of a stat |

s2 | variance of a sample |

sd or s or omega | standard deviation of a sample |

Median when is it applicable? | distribution is skewed,less sensitive to extreme scores |

(+) (-) distributions where do the scores fall? what are they called? | (+)-mean, median mode - -mode median mean measures of central tendency |

Variability | spread of scores |

Most common measure of variability | standard deviation |

% in a normal distribution | 34.1, 13.6, 2.1 (both sides) |

Positive skewed, and negative skewed | positive tail in the back, negative tail in the front |

statistical significance | probability that the result of the study could have occurred by chance |

0.05 p is what? | level of significance standard, 5% risk not a true difference, 95% true |

T-test broad | Measures difference between: 2 groups at a single time or one group at two time point |

Dependent t test | Differences within a group @ two time points on a single dependent variable, differences between a group on two different dep. variables |

Independent t-test | between-difference in mean scores between two groups, unrelated to each other, at one point, single dep. variable |

df | degrees of freedom or number of values in the study free to vary based on the # of participants in the study |

one way ANOVA | compare differences between 3 or more groups a single dep. measure similar to ind. t-test |

Repeated Measures ANOVA | Compare the differences of one group 3 or more times (dep. t -test) |

Mixed model ANOVA | Both between group and within group, differences simultaneously providing an interaction effect, provide separate results- between group differences and within groups diff(main) |

Analysis of covariance ANCOVA | Compare differences between and/or within groups while statistically controlling a variable (covariable) |

ANOVA | more than two means compared |

Correlations | degree at which two variables flucuate together |

p<0.05 p>0.05 | interaction effect exists no interaction effect performed similiar |

r ranges what does it mean if + or - | 0-1.0, +- related in the same direction speed and distance -- - related in opposite direction, speed and time- faster speed less time |

PPMC | Inferential stat that examines the strength of relationship between two continous variables |

Spearman | Same as PPMC but compares variables that are rank ordered |

Odds ratio | OR= AD/BC-odds of presence/absence of one variable is associated with the presence/absence of another variable |

Linear regression | identify predictors of continuous outcome measure (ROM- numbers mean something) |

Logistic regression | identify predictors of categorical outcome measure (only two outcomes) |

Factors that have the greatest impact on a particular outcome | predictors |

Total variance | r2 (correlation squared) 100- r2 is the variance unaccounted for variance is always smaller than correlation |

CI | reliability- range of outcomes expected |

Difference between clinical and stat. significance | stat. is based on sample size- likely to find a difference with a large sample size, could have a small mean difference (1.5) but that is not clinically significant |

Cohen's d | effect size measure, difference between two groups in standard deviation measure |

R is effect size in relationship studies, what does it mean if r value is 0.87 vs. 0.2 | 0.87 greater effect, stronger relationship 0.20 smaller effect, lesser relationship |