Core Knowledge 5% Flashcards
“Male” or “female” are examples of what type of variable?
Categorical or nominal variable.
For categorical variables, each subject can be placed into one of the categories. If there are only 2 possible outcomes, it is a dichotomous variable.
Severity score of swelling (0 to 3+) is an example of what type of variable?
Ordinal variable.
Ordinal variables have an inherent order. Another example would be tumor staging. Order exists among the categories, but the difference between adjacent categories is not uniform.
Auditory ossicles are part of which section of the skeleton, axial or appendicular?
Axial
The ribs are part of which section of the skeleton, axial or appendicular?
Axial
The musculoskeletal system comes from what embryonic layer?
Mesoderm, appears at 4th week of gestation
If a study is “reliable,” what does that mean?
Consistency of an instrument or measurement. The measures devised for concepts are stable on different occasions.
If a study is “valid,” what does that mean?
Validity refers to the accuracy of an instrument or measurement.
What is the concept where “the findings can be generalized to other social settings”?
External validity
Match the animal model up with the disease it models:
a. UCD-200 chickens
b. MRL/lpr mice
c. Injection of type II collagen with adjuvant into mice
d. Neonatal mice infected with coxsackievirus (picornaviruses)
a. UCD-200 chickens - tight-skinned chickens, scleroderma
b. MRL/lpr mice - These mice have a mutation in Fas gene affecting apoptosis. These mice develop lupus-like phenotype including glomerulonephritis. Other potentially useful models of lupus including the New Zealand black or white mice, BXSB, MRL-gld, and various knockout mice.
c. Injection of type II collagen with adjuvant into mice (and also various adjuvants, and streptococcal cell wall antigens) - arthritis
d. Neonatal mice infected with coxsackievirus (picornaviruses) - myositis
What type error occurs when the investigator falsely rejects the null hypothesis, concluding that the intervention has an effect that is not present?
Type 1 error
A type I error occurs when the investigator falsely rejects the null hypothesis, concluding that the intervention has an effect that is not present.
What type error occurs when the investigator falsely accepts the null hypothesis, concluding that the intervention does not have an effect, when an effect is indeed present
type II error
A type II error occurs when the investigator falsely accepts the null hypothesis, concluding that the intervention does not have an effect, when an effect is indeed present
What is reporting bias?
Reporting bias refers to a lack of objectivity in dissemination of research results in the literature.
For example, positive studies in which the null hypothesis is rejected are much more likely to be published than negative studies. In addition, study sponsors with a stake in the outcome (eg, pharmaceutical companies sponsoring a drug trial) can attempt to influence publication of the results.
What is response bias?
Response bias occurs when subjects give the answer they believe to be desired by the investigator, or the response they believe is consistent with the expected outcome of the research. Response bias can be either conscious or subconscious.
What is Propensity scoring?
Propensity scoring attempts to quantitate how similar patients are at baseline in comparison with the treatment used.
It establishes a likelihood that a patient will receive a treatment based on baseline variables. Each subject then receives a propensity score and is matched to similar patients, and outcomes are compared among treatment groups, controlling for the score.
The use of propensity scoring minimizes confounding by indication. Matching and stratification by disease severity factors are other methods to help mitigate confounding in observational cohort or case-control studies where blinded randomization is not possible.
Which statistical test compares categorical variables in a sample to see if they are related (eg, to see if the distributions differ from one another).
A chi-square (χ2) analysis (also known as the χ2 test for independence)
The χ2 test compares variables in a sample to see if they are related (eg, to see if the distributions differ from one another). A low χ2 value indicates a high correlation between the sets of data. If the χ2 is zero, then the observed and expected results would be completely equal. Whether a χ2 is high enough to be statistically significant can be determined from the P value (usually P < 0.05 is considered statistically significant).
