click below

click below

Normal Size Small Size show me how

# Statistics 1

### UWORLD + FA Statistics Review

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

Correlation coefficients range from ____ to ____, and are used to describe ________________ and ________________ of a linear association. The ________________ to ___ or ____, the ______________ the association. | -1 to +1; strength and polarity; Closer to -1 or +1; STRONGER |

What kind of cancer is the most common in men and women? | Lung cancer |

What is the most important factor for lung cancer development? | Tobacco use |

Order of female cancer mortality | Lung> Breast> colon> Pancreas> stomach |

Sensitivity is? | Refers to the ability of the test to identify those with the disease. |

Screening for a life-threatening diseases, prefer a __________ ________________ test. | High sensitive |

A sensitive test yields a ________ number of _______________ which helps to __________________. | High number of False positives; Rule OUT |

What is the reason of an SPECIFIC test? | It is the ability to identify those without the disease |

Specific test yield a ________ number of __________________, which helps to ____________________. | Low number of False positives; Rule IN |

Low false-negative is related to | Sensitivity |

A person who has a disease or might have it test positive | Sensitivity |

Sensitivity | Rule out |

Specificity | Rule in |

Sensitivity equation | = TP /(TP + FN) |

Proportion of all people without the disease who test negative, or the probability when the disease is absent, the test is negative. | Specificity |

Low-false positive rate | Specificity |

______ ____________ test is used after a positive screening test | High Speficific |

Specificity equation: | = TN/(TN + FP) |

PPV | Positive Predictive value |

Probability of a person that test positive actually has the disease | PPV |

PPV varies directly with: | Pretest probability |

High pretest probability = | High PPV |

NPV varies _________________ with ______________ and ____________. | INVERSEBLY; pretest probability and Prevalence |

Case-fatality rate is a measure of? | Severity of a condition |

Number of DEAD by condition divided by number of people alive with the condition. | Case-fatality rate |

Ignorance of the test accords | Blinding technique |

A blinding technique reduces the: | Observer bias |

What is recall bias? | Study participants inaccurate recall of post exposure and happens in most POST-EXPOSURE studies such as case-control studies. |

What kind of study is used in order to reduce recall bias? | Prospective study |

Study done in order to record the initial exposure status and have tight control | Prospective study |

____________________ bias is NOT representative of the general population. | Selection bias |

What are the Observational studies? | Cohort (retrospective and prospective), case-control, cross-sectional, and ecological studies |

Cohort studies are also referred as _________________ studies. | Longitudinal |

What are the subcategories of Cohort studies? | Retrospective and Prospective studies |

Retrospective studies are affected by: | Recall bias |

What value/unit answers the question, 'What is the probability that a patient has a disease?'? | PPV |

What is the main determination of the Cutoff value? | Measure whether a result is interpreted as positive or negative |

Lowering cut off value: | - moves midline to the LEFT - Increase False Positives and Sensitivity - Decrease False Negatives and Specificity - Increase NPV and Decrease PPV |

What happens with an increase in cutoff value? | - moves midline to the RIGHT - Increases false negatives, Specificity, and PPV - Lowers false positives, Sensitivity, and NPV |

What is the equation for Power? | = 1-B |

Power: | Measures the probability of rejecting a null hypothesis when it is truly false |

What are the 2 important factors that determine Power? | 1. Sample size 2. Difference between outcomes |

Power normal value is usually at | 80% |

B-error | probability of making a type II error |

Investigated to reject a null hypothesis, when it is really true | Type I error |

What kind of error is encountered in when investigator find a statistical difference where there is none? | Type I error |

What represents the maximum acceptable degree of type I error? | a (alpha) |

a(alpha) is usually compared to what other value? | p-value |

What is the usual value of "p-value"? | 0.05 or 5% |

Null hypothesis (H0); | Hypothesis of no difference or relationship |

Stating there is not a difference or relation where there really is one | Type II error |

B(beta): | probability of making a type II error |

B(beta) is related to ___________________________. | Statistical power (1-B) |

What is statistical power? | The chance of rejecting a null hypothesis when it is false |

What are the 3 factors that increase power and decrease B? | 1. Increase sample size 2. Increase expected effect size 3. Increase precision of measurement |

False-positive error is another way to refer to _______________. | Type I error |

Another way to refer to a Type II error? | False- negative error |

A 95% Confidence Interval (CI) corresponds to a _____________, which the most often used values. | a(alpha) = 0.05 |

Range of values within the true mean of the population is expected to fall, with a specified probability? | Confidence Interval (CI) |

What is the most commonly used CI? | 95% |

What is the value of "Z" in a 95% CI? | 1.96 |

What is the value of "Z" in a 99% CI? | 2.58 |

If the 95% CI for 2 values includes 0: | 1. There is NO significant difference 2. H0 (null hypothesis) is NOT rejected |

If the CIs of 2 groups do not overlap, it means? | There is a statistical difference |

What is the meaning if the CIs of 2 groups overlaps? | There is NO statistical difference |

If the 95% CI of odds ratio and relative risk includes ______, then _____________________________. | 1; H0(null hypothesis) is not rejected |

What is the equation of chance is the probability is set at x and there is y number of samples? | = x (to power of y); multiply x itself y number of times. |

What is determined by a Case-control study? | The EXPOSURE STATUS of people with disease (cases) to the exposure status of people without the disease (control). |

What it is the main association of Case-control study? | Odds ratio (OR) |

What is the equation for odds ratio? | = (odds of Exposure in cases)/ (odds of Exposure in control) |

= ad/bc | Odds ratio (OR) |

Values (largest to smallest) in a NEGATIVELY skewed distribution . | MEAN>>Median>> mode |

Mean, median, and mode are arranged as __________________________________________ in a POSITIVELY skewed distrugion. | mode> Median> MEAN |

In a graphic depiction of a negatively skewed distribution, the graph "appears" to be ______________________ to the __________. | PUSHED to the RIGHT |

A positively skewed graph, appears to be ___________ to the ______. | PULLED to the LEFT |

In a normal (Gaussian) distribution, the normal SE account for the following percentages: | 1 SE --> 68% 2 SEs ----> 95% 3 SEs --------> 99.7% |

ANOVA: | Analysis of variance |

ANOVA is used to determine any significant differences between the _____________________ of ________________________. | Means; 3 or more groups |

A t-test is used in analysis of ______ groups. | 2 |

What is the overall advantage of a Meta analysis? | Gathers data of several studies, leading to greater statistical power. |

What testo of analysis would be used to compare the blood pressure between men and women? | t-test |

Checks differences between 2 or more percentages or proportions of categorical outcomes | Chi-square test |

Chi-square does NOT measure the ________ values. | Mean |

In labeling a 2x2 square values for calculating ODDS RATIO, what are the description of the following: | A+C --> total people with disease B+D --> total people without disease A+B --> total people with Exposure C+D --> toal people without Exposure |

What is the attack rate: | a ratio between people who contracted a disease divided by the people who is at risk of contracting the disease. |

3 standard deviation account for ________% of the sample results. | 99.7% |

A result with 2 standard deviations is within what percentage of all the sample population? | 95% |

68% of all population in a result is withing how many standard deviation? | 68% |

What is risk? | Probability of developing a disease over a certain period of time |

How is risk calculated? | = (number of affected subjects)/ (number of corresponding subjects in that group) |

Relative Risk (RR) is usually a measure seen in ________________ studies. | Cohort |

Risk of developing disease in exposed group divided by risk of unexposed group | Relative Risk (RR) |

In what cases or situations OR approximates to RR? | In rare diseases or those with low prevalence |

What does an RR=1 means? | No relation between the exposure and disease |

RR < 1 means: | Exposure associated with decreased disease occurence |

RR of 1.5 would signify that: | Exposures associated with INCREASED disease occurrence |

RR is calculated by: | = a/(a+b) / c/(c+d) |

Difference in risk between the exposed and unexposed groups | Attributable risk (AR) |

= a/(a+b) - c/(c+d) | Attributable risk (AR) |

Risk of those in the exposed group is calculated by: | = a / (a+b) |

What would be the AR if lung cancer in smokers is 25% and lung cancer in nonsmokers is 10%? | 15% |

The proportion of risk reduction attributable to the intervention as compared to a control. | Relative Risk Reduction (RRR) |

If 3% of people who receive their annual flu vaccination developed the disease, and 12% of those who decided to remain unvaccinated developed the disease. What is the RR? and RRR? | RR is 3/12 = .25 or 25%, and RRR is 1-RR, so, RRR=1-.25= 0.75 |

How is relative risk reduction (RRR) calculated: | = 1 - RR |

The difference in risk (not the proportion) attributable to the intervention as compared to a control | Absolute risk reduction (ARR) |

While AR is the difference (subtraction) of the risk of EXPOSED group minus the risk of those UNEXPOSED group, in ARR: | The subtraction reverses the values of the subtraction, having the risk of UNEXPOSED - risk of EXPOSED |

ARR is caluated: | = C/(C+D) -- A(A+B) |

NNT: | Number of patients who need to be treated for 1 patient to benefit |

What value is needed in order to calculate for NNT? | ARR |

What is the equation for NNT? | = 1/(ARR) |

One divided by the difference in risk (not proportion) attributable to the intervention as compared to a control. | Number needed to treat (NNT) |

What does it mean for the treatment quality, a low NNT? | The lower the NNT, the better the treatment |

Number of patients who need to be exposed to a risk factor for 1 patient to be harmed | Number needed to Harm (NNH) |

What value of risk is needed to calculate NNH? | Attributable risk (AR) |

What is the equation to calculate NNH? | = 1/ (AR) |

A high NNH number means: | Safer exposure |

The consistency and reproducibility of a test | Precision |

Another term used to refer to "PRECISION" | Reliability |

Trueness of test measurements accounts or describes _______________. | Accuracy |

Matching in a study is used for: | Control confounding matching variables in a case-control study. |

What are the potential (common) confounding factors associated with matching? | Race, age, neighborhood, diet |

Case control studies desire _____________________ variables, in order to approximate a similar distribution. | Matching confounding |

What is a "Prevalence" or a "Snapchat" study? | Cross-sectional study |

What is measured or researched in a Cross-Sectional Study? | Exposure and outcome simultaneously of a particular POINT in TIME. |

What kind of study is designed to answer the question, "What's happening?" | Cross-sectional study |

What is the MC measure appointed in a Cross-sectional study? | Disease Prevalance |

Frequency of disease and frequency of risk-related factors are assessed in the present | Cross-sectional study |

Cross-sectional study can show ______________________, but it does not establish _____________________. | Risk factor association with disease; Casuality |

A Case-control study is designed to answer what question? | What happened? |

Case-control study: | Compares a group of people with disease to a group without disease |

What is the value obtained in designing a Case-control study? | Odds ratio |

Odds ratio is often calculated in what kind of study? | Case-control study |

A cohort study answers the questions: | 1. Who will develop disease? (Prospective) or, 2. Who developed the disease? (Retrospective) |

What kind of study is designed to look to see if exposure or risk factor is associated with later development of disease. | Cohort study |

Compares a group with a given exposure or risk factor to a group without such exposure. | Cohort study |

Relative Risk (RR) is a measurement obtained in what kind of study? | Cohort study |

Study Design --------->Measurement: Cohort Study ------------------------------> Case-control study -----------------------> Cross - sectional study -------------------> | Relative risk Odds Ratio Disease Prevalence |

What bias is involved in recruiting of subjects? | Selection bias |

Methods used to reduce Selection bias? | 1. Randomization 2. Ensure the choice of the right comparison/reference group |

What is Berkson bias? | A type of Selection bias, in which the selected population of a hospital is less healthy than the general population |

What is the most common error leading to a Selection bias? | Sampling bias |

Nonrandom sampling or treatment allocation of subjects such that study population is not representative of target population | Selection bias |

What bias or errors are committed while performing the study? | 1. Recall bias 2. Measurement bias 3. Procedure bias 4. Observer-expectancy bias |

1. Confounding bias 2. Lead-time bias 3. Length-time bias | Errors or bias that occur while interpreting the results |

What is the most common bias seen in retrospective studies? | Recall bias |

Subject's awareness of disorder, is a predisposition to: | Recall bias |

A patient recently noticed a previous exposure to a specific risk factor after reading similar case in the newspaper. | Recall bias |

How is recall bias reduced? | Decrease time from exposure to follow-up |

What is the Hawthorne effect, and what does it lead up to? | Participants behavior changes upon awareness of been observed, leading to Measurement bias |

A faulty blood pressure cuff will lead to a ____________________. | Measurement bias |

A placebo group would reduce the ____________________ | Measurement bias |

Information is gathered in a distorted manner, either due to equipment malfunction or by odd or changing behavior of subject. | Measurement bias |

Pygmalion effect is: | The belief of the researcher efficacy of the treatment, which results in changes in outcome of that treatment. |

A physician with a daughter with terminal cancer, is in charge of research a new chemotherapeutic drug. This represents what kind of challenge to the study? | Observer-expectancy bias may result due to strong need or desire of the physician. |

An observer with strong desire of favorable outcomes is more likely to notice positive results. | Observer-expectancy bias |

Subjects in different groups are not treated the same | Procedure bias |

What are 3 methods or procedures used to reduce Confounding bias? | 1. Multiple/Repeated studies 2. Crossover studies 3. Matching |

Subject acts as their own controls | Crossover study |

Study in which participant is subjected to several sequence of treatments or exposures | Crossover study |

Patients with similar characteristics in both treatment and control group | Matching |

Confounding bias | When a factor is related to both the exposure and outcome, but not on the causal pathway, it distorts or confuses effect of exposure on outcome |

A wash-out period is seen in what kind of study design? | Crossover study |

If variable affects both, the exposure and outcome, is known as __________________ bias. | Confounding bias |

Confounding: | A situation in which the effect or association between an exposure and outcome is distorted by the presence of another variable |

When early detection is confused with increased survival | Lead-time bias |

Length-time bias: | Overestimation of survival duration due to the relative excess of cases detected that are slowly progressing |

Lead-time bias: | Overestimation of survival duration due to earlier detection by screening than clinical presentation |

In a slowly progressive cases of lung cancer, there is a chance of________________, due to ____________________________. | Length-time bias; the slow progression of condition leads to a increased number of cases. |

A quick and effective screening test for HIV infection, is a risk to the development of what kind of bias? | Lead-time bias as it increases the "apparent " survival duration, but not the overall mortality rate. |

Participants without the disease in a study. | Control group |

Previous exposures status is evaluated by _________ study. | Case-control study |

= Control rate - Case (treatment) rate | ARR |

= ARR/ (control rate) | RRR |

= (Treatment rate) / (control rate) | RR |

What is a case-series study? | A descriptive study that tracts patients with a known condition to docucient the natural history or response to a treatment. It provides no statistical significance measurements |

Most frequent value in a data set | Mode |

Value located in the precise center of an ordered data set | Median |

Average of all data set? | Mean |

In a positively skewed distribution, the greatest measurement of the data set is the _____________. | Mode |

What is meta analysis? | The result of several trials to increase statistical power and provide overall pooled effect estimate |

The null value is dependent on the _________. | Confidence Interval (CI) |

If the CI DOES NOT include the null value? | Result is statistically significant |

If the study or case is to reflect a DIFFERENCE in values, the null value is _____________. If the CI in this case does NOT include the _____, the results are __________________________. | 0; 0; statistically significant |

= (Absolute risk control) - (Absolute risk treatment) / (Absolute risk control) | RRR |