Biology simple sampling

1. Problem statement: articulates the problem and an argument that explains the need for the study. 2.Purpose statement:aim of the study. 4.Independent variables: Presumed Cause/influence,Will Cause change in outcome of interest,May Be an intervention Dependent Variable: presumed effect, outcome variable,variable researchers are trying to understant, explain, predict. 7.Experimental: RCT, QUASI,Nonexp(observational): decriptive, Cross-sectional,cohort, case control, 8. RCT a:Best Evidence that the intervention caused the outcome,Gold Standard;Causation;Multiple RCTs can become Meta-analysis D:Many Research variables cannot be manipulated (either realistically or ethically;May Not be possible to randomize into group; Subject To the Hawthorne Effect; Quasi A:Enhanced Practicality;Some Control;Participants Are not always willing to be randomized D:Cause And effect hampered (something else may have affected the change in the DV);Non Equivalence between groups; NONEXP A:Studies Problems that cannot be conducted experimentally;Can Gather information on relationships between multiple variables;Experimental Studies are often dependent on starting with descriptive correlational.D:No Causation;Self-selection – groups within a study that form themselves (male vs. Female; smokers vs. Non Smokers) 9. Manipulation-is there an intervention Observation- when are the researchers collecting data, how often Randomization Control - how many groups are being compared.10.Counterfactual-What is done to the control group instead of the intervention in exp and quasi exp: no intervention, no treatment, Alternative treatment, Placebo,Standard care or usual care, Modified treatment (lower dose , partial intervention), delayed treatment(waitlisting control group). 12. Probability - (random)Each Element of an accessible populationhas an equal independent chance of being Selected to the sample.>Simple Random Samplin Requires a sampling frame List of all available members in the Population E.G. Telephone book All students at RRC>Stratified Random Sampling The sampling frame is divided into 2 (or More) list based on strata and random sampling is completed from each list Separately Sampling via homogenous subsets (proportional to population demographics) Proportionate sample Disproportionate sample – for comparison Studies>Cluster Sampling Successive random sampling –multistage Sampling>Systematic Sampling Every kth case – 10th, 5th, 3rd, 2nd The kth number is known as the “sampling interval”>NonProbability-(non random)>Rarely Representative of the target population (i.E. Much larger sampling bias/error Will result)>Elements Do not have an equal chance of selection in the sample and often are referred To as “self-selected>Non-representative – readily available individuals are rarely typical of a population>Convenience uses most available (everyone in target Population approached if target pop. Is small) Volunteers Newspaper ads or poster and pamphlet Campaigns Snowball” -- network/chain sampling/word Of mouth Useful as a starting way to obtain a Sample when little is known about a topic>Consecutive Sample:  Recruitment of all people from An accessible population over time -- in situations where names are Continuously added to a list or new cases are diagnosed on a regular basis (“rolling enrollment”) E.G. Surgical waiting lists, new pregnancies added to a obstetrics team patient List, new hip fractures, new bed sores. >Quota proportional representation of population Strata(e.G. Gender; ethnicity) Works well in consecutive sample Scenarios where you are likely to get an underrepresentation of a particular Strata by convenience alone.>Purposive matching a comparison group in a Quasi-experimental study Other purposeful choosing of experts on a Topic.14.Predicts the appropriate sample size ex The power analysis predicted an ideal sample for this study, to Make the statistics be as reliable as possible (and reduce the risk of poor Statistical conclusion validity), would be 89 participants. 16.Informed Consent>letter signed by participants, implied consent(return of the questionnaire,Process consent(ongoing qualitative); Confidentiality Procedures>anonymity number rather than names, Securing information>Debreifing and Referals,sharing findings,venting sessions,reffering to health social psychological services.18. T-test – used to test the statistical significance of a difference between means of Two groups (t) Parametric (assumption of a normal curve) Independent group t-test (means of different groups) Dependent group t-test or paired t-test (means of same group E.G. Before-after)One tailed or two tailed>ANOVA – Analysis of Variance – measures significant differences in means of three or More groups (F) Parametric (assumption of a normal curve) F-ratio Multiple comparison procedures (post hoc Tests) – to find out which group caused the significant difference Repeated measures ANOVA – when means are Compared at different points in time (pretest to posttest 1 and posttest 2)>Chi-squared Test (χ2) Non-parametric test (does not require a Normal curve) Compares the differences in proportions In 2 or more groups, with variables with a small number of categories, Nominal data/ordinal, 2 x 2 crosstab table, Friedmans also an (χ2) used with larger category nominal/ordinal.>Correlation Coefficients, Pearsons(r),parametric, r=magnitude and direction of a relationship between two variables, Any correlation greater than r=0.70 is considered strong. >Multivariate SA deals with comparing three or more variables simultaneously.>Multiple regression uses F tables when squared R2.>Ancova combo of anova and multiple regression F statistics.>Manova Multivariate and anova