Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
24 Cards in this Set
- Front
- Back
|
SP01 [Mar96] [Jul98] [Jul99] [Feb06] Tests that use ranking of data: |
ANSWER D
A. Can be applied to any distribution : FALSE they cannot fit normal distributions or t-tests as they are not continuous B. Include the chi square test : FALSE used in categorial nominal C. Have greater power than non-ranking tests : FALSE Non-parametric has less power (ei will fail to reject Ho when there is a difference) D. Are preferred when normal distribution cannot be confirmed : TRUE usually when sample is small and normal distribution cannot be demonstrated |
|
|
SP02 [Mar96] [Feb00] [Jul02] [Mar03] [Feb06] Standard error of the mean:
A. Is proportional to N B. Is greater for sample than SD of population C. Measures variance within a sample D. Measures dispersion around population mean E. The difference between the sample mean and the population mean. |
ANSWER D
Standard Error of the Mean -is not the variance of the sample -it is the variance of the sample means (ie btn difference populations) -SEM=sd/√n -can be used to define the 'range' for the true population mean Therefore SEM is inversely proportional to √n |
|
|
Standard error of the mean:
A. Measure of sample variability B. Measure of difference between sample & population mean C. SEM is always less than SD D. measure of the variability of sample means E. SEM is always greater than SD |
ANSWER D
Standard deviation (SD) of a sample is one measure which describes the variability within a single sample. Another measure of dispersion is the variance which is equal to the square of the standard deviation. Standard error of the mean (SEM) is the standard deviation of the "sampling distribution" (& not the "sample distribution"). An estimate of its value can be calculated from the sample SD; it describes variability of the means from all possible samples of a particular sample size (n) from the mean of the "sampling distribution" (which is useful because this mean is equal to the population mean). |
|
|
SP03 [Mar96] [Jul98] [Mar02] [Jul02] Use of chi-square test inaccurate with:
A. 2x2 contingency table B. Expected value of any cell < 5 C. Observed value in any cell < 5 D. Sample is not normally distributed E. Measured variables are independent |
Answer - B
(Expected value in any cell <5) and B (Use Fisher test if observed <5) Chi X2 test -non parametric test : categorial nominal data -uses contingency table Rules 1. The sample must be randomly drawn from the population. 2. Data must be reported in raw frequencies (not percentages); 3. Measured variables must be independent; 4. Values/categories on independent and dependent variables must be mutually exclusive and exhaustive; 5. N should be >20 and Expected>5 Can use Yates Correction if expected<5 |
|
|
Chi square contingency tables
A. Use Fisher test if observed <5 B. Use Fisher test if expected <5 C. Use Fisher test if N>50 D. Use Fisher test if greater than 2 x 2 contingency table E. Parametric data E. |
ANSWER B
The chi-square statistic becomes inaccurate when used to analyze contingency tables that contain exactly two rows and two columns, and that contain less than 50 cases. Fisher's exact probability is not plagued by inaccuracies due to small N's. Therefore, it should be used for two-by-two contingency tables that contain fewer than 50 cases. |
|
|
SP04 [Jul97] The mean in a very large sample:
A. Numerically greater than the standard deviation B. Is always equal to the mode C. Is more than the median D. Represents a normal distribution E. Gets larger as the sample size increases |
ANSWER A
A : as n increases, Sd decreases B :the mode is the most common variable, mean is the average and will be affected by skew and outliers to a point C: the median is middle value D: the distribution of the mean will be normal as n increases E: gets smaller |
|
|
SP05 [Mar98] [Jul02] [Mar03] The standard normal distribution:
A. Standard deviation equals 68% B. Mean, median & mode are the same C. Mean is one D. Mode is one E. Median is one |
ANSWER B
A. Incorrect. 68% probability of that the sampled variable will sit within 1 SD of the mean. SD should equal 1 B - correct as mean, median and mode the same with a standard distribution C - Incorrect; mean is zero in a standard normal distribution D - Incorrect; mode is zero E - Incorrect; median is zero |
|
|
ST06 [Mar98] [Jul00] In a study for depth of epidural catheter insertion, the mean is 4.4 and the standard deviation is 0.3 Which ONE of the following is true?
A. If a normal distribution, 68% of values wold lie between 4.1 and 4.7cm B. None was greater than 5.5 cm (or ?6.8cm) C. The least distance was...?? D. 99% of the sample lies within 1.96 SD of the mean E: 500 patients had catheters at some length. |
ANSWER A
|
|
|
SP07 [Mar98] Simple linear regression:
{graph of straight line crossing y axis at +3} A. y = 3 + 6x B. y = 3 + 0.6x C. ? D. None of the above |
Both A and B correct
Need to see the graph to determine the gradient of the line. However, with the given information both A and B are true. * For x=0, y=3; for both. In addition, the gradient of the straight line will be the "rise over run" For a simple linear function... y = mx + c |
|
|
SP08 [Jul98] [Mar99] [Jul01] Which one of the following statements regarding the standard deviation is true?
A. Mean +/- one SD includes 50% of values B. Mean +/- one SD includes 66.7% of values C. Mean +/- two SDs include 99% D. Mean +/- three SDs include 99.73% E. Mean +/- 1.96 SD includes 99.73% |
ANSWER D
For a normally distributed population: * mean +/- 1 SD includes 68% of values, * mean +/- 2 SD includes 95.4% of values, * mean +/- 3 SD includes 99.7% of values and mean +/- 1.96 SD includes 95% of values |
|
|
SP09 [Jul98] Ordinal data:
A. Assumes a normal distribution B. Includes non categorical data C. Weight in kgs D. Parametic tests E. Mann-whitney U |
ANSWER E
Ordinal Data -Categorical data -has an order/rank eg Pain Scores -but have no obvious correlation with each other (ei pain score of 8 is not 4 times more than pain score of 2, we can only say it was more painful) -they not NOT assume normal distribution Non Parametric tests are applied Ordinal Tests include Single comparisions -Mann-Whitney U -Kolmogorov-Simimmov -Wilcoxin rank test Multiple comparisons -Kruskal-Wallis -Friedman rank |
|
|
SP10 [Feb00] Paired t-test
A. Assumes the normal distribution B. Is a non-parametic test C. two samples are dependent D. Unpaired t-tests has greater power than paired t-tests E. Measures the variance between two samples |
ANSWER A
Student's T-test -parametric test Assumptions -samples sizes are equal -samples have the same variance -samples are normally distributed -samples are independent -population must be normally distributed -or sample must be manipulated to become normal (by taking logs or inverse) Paired has greater power than unpaired t-test |
|
|
SP11 [iq In a clinical trial, a patient either vomits or not. What type of data is this?
A. Ordinal B. Nominal C. Ratio D. Interval |
ANSWER B
|
|
|
SP12 [Feb00] [Jul00] [Apr01] Odds ratio:
A. Is prevalence vs. incidence B. Gives an indication of ?? in exposed vs non-exposed patients C. Formula is Number of positive outcomes/ Number of negative outcomes D. Gives the prediction of a disease outcome knowing the risk factors E. Gives prediction of risk factors with a known disease outcome |
ANSWER C
A. is incorrect B. the question may be 'gives an indication of risk in exposed vs non-exposed patients' which would be correct C. the odds of an event are calculated by 'positive' outcomes/'negative' outcomes. the odds ratio is calculated by dividing the odds in the exposed group by the odds in the control group D. i'm unsure what this answer is getting at E. and unsure about this one too |
|
|
SP13 [Jul00] With respect to 95% confidence intervals:
A. Equals mean +/- 1.96 SE B. Will contain the population mean 95% of the time C. Tells variability of sample D. Tells 5% chance of finding sample result E. Assumes a normal distribution |
ANSWER B
A. False : B. True C. False- this is standard deviation D. False E. False Confidence intervals are calculated from the standard error of the mean, which is calculated from the standard deviation and sample size. The confidence interval is a range, with a 95% probability that the population value lies somewhere within this range. The confidence interval can be calculated for a number of different things, including mean, risk ratio, median, and regression analyses. Confidence intervals allow estimation of population parameters from sample statistics. The CI gives an estimate not only of the parameter itself, but also of precision. The wider the CI, the less precise the value derived from the sample. |
|
|
SP14 [Jul01] [Jul04] Students t-test
A. Used to compare 2 groups B. Used if groups have different variance C. For small size samples D. ? E. ? |
ANSWER A
A. True - Analysis of more than two groups shoud be done with ANOVA (or using the t-test and a Boferroni correction - but this is not recommended). B. False - one of the assumptions is that the samples have the same variance C. False - if small sample size, use non-parametric tests |
|
|
SP15 [Jul01] [Jul04] All of the following tests EXCEPT one, can all be used to compare two dissimilar groups:
A. Chi square B. Mann whitney U test C. Wilcoxon signed ranks sum test D. Spearman rank order E. Kruskall Wallis |
ANSWER D
A - Incorrect B - Incorrect C - incorrect D - correct because while the others look for a difference (albeit sometimes by trying to prove the null hypothesis - that they are the same - is incorrect), the Spearman rank order looks for a strength of correlation E - incorrect |
|
|
SP16 [Feb04] The central limit theorem states that:
A. Long option mentioning mean, median and mode B. Best measure of central tendency is mean C. With repeated sampling, distribution approaches that of normal distribution D. With increasing sample size the sample means approximate a normal distribution. E. If the 95% confidence interval includes zero then it is not statistically significant |
The central limit theorem states that, "Regardless of the shape of the frequency distribution of observations of the original population, the frequency distribution of sample means of repeat random samples of size n tends to become normal as n increases."
NOTE: this is a poorly remembered question A. ? B. the best measure of central tendency is mean, in a normal distribution. in a skewed distribution, the median may be a better measure. C. with repeated sampling (eg n>100), distribution approaches that of normal distribution. This is true even if the distribution in the population the sample is taken from is not normal. D. is the best answer - "The distribution of an average will tend to be Normal as the sample size increases, regardless of the distribution from which the average is taken except when the moments of the parent distribution do not exist. All practical distributions in statistical engineering have defined moments, and thus the CLT applies."[1] (http://www.statisticalengineering.com/central_limit_theorem.htm) E. if the 95% confidence interval includes zero then it may not be statistically significant, but it depends on the data being collected. |
|
|
SP17 [Feb04] [Jul04] What kind of data is the ASA classification?
A. Nominal. B. Ordinal C. Interval D. Ratio. E. Parametric. |
ANSWER B
|
|
|
SP18 [Feb04] [Jul04] Repeated statistical testing:
A. Increases alpha (Type I) error B. Increases beta (Type II) error C. Decreases power D. ? |
A is correct
Repeated testing increases the likelihood of finding a difference due to chance alone, and incorrectly rejecting the null hypothesis. Thus it increases the chance of a type I error. |
|
|
The 95% confidence interval for a proportion:
A. Cannot be calculated if the sample size is small. B. Is the interval within which 95% of sample proportions would lie if we were to take repeated samples of a given size from the population. C. Is the interval within which we expect the population proportion to lie with 95% certainty. D. Is wider than the 99% confidence interval for the proportion. E. Is calculated as the sample proportion ± standard error of the proportion. |
ANSWER B
A. FALSE. A confidence interval can be calculated whether the sample size is large, using the Normal approximation to the Binomial distribution, or small, using the exact Binomial distribution. It will be very wide if the sample size is small. B. TRUE. C. FALSE. the confidence interval is usually interpreted in this way. D. FALSE. The 95% confidence interval is narrower than the 99% confidence interval. We need to be more certain that the 99% CI will contain the true proportion than if we have a 95% CI, and so the CI is wider. E. FALSE. The 95% confidence interval is calculated as the sample proportion ± 1.96 times the standard error of the proportion. The 1.96 is often approximated by 2. The sample proportion ± standard error of the proportion is the 67% confidence interval for the proportion. |
|
|
In a prospective study of post-traumatic stress disorder in children involved in road traffic accidents (Stallard P et al. BMJ 1998; 317: 1619-1623), post-traumatic stress was found in 41/119 children (34.5%, 95% CI 25.91% to 42.99%).
Select the statement which you believe to be true: A. In the population of children involved in road traffic accidents, the true rate of post-traumatic stress is greater than 34.5%. B. In this sample, the rate of post-traumatic stress could take any value between 25.91% and 42.99%. C. The true rate of post-traumatic stress in the population of children involved in road traffic accidents is unlikely to be less than 25.91%. D. The authors would have obtained a more precise estimate of the proportion of children with post-traumatic stress if they had calculated a 99% confidence interval. E. If an earlier study into the same issue had reported that the rate of post-traumatic stress in children after road traffic accidents was only 27.0%, we could conclude that the rate of post-traumatic stress was increasing. |
ANSWER C
A. false. In the population of children involved in road traffic accidents, the true rate of post-traumatic stress is likely to lie between 25.91% and 42.99%. B. false. The confidence interval relates to the population not the sample. In this sample, 34.5% of children experienced post-traumatic stress. C. true. There is only a 2.5% chance that the true rate of post-traumatic stress in the population of children involved in road traffic accidents is less than 25.91%. D. false. The authors would obtain a more precise estimate of the rate of post-traumatic stress if they had taken a larger sample size. Increasing the width of the confidence interval simply increases our confidence that the true rate of post-traumatic stress lies in the interval. E. false. This may indeed be the case. However, as the confidence interval for the current study includes 27.0%, we cannot rule out the possibility that the true rate of post-traumatic stress equals 27.0% in the population. |
|
|
A Type II error is also known as
A. False positive B. False negative C. Double negative D. Positive negative |
ANSWER B
|
|
|
37. This is the difference between a sample statistic and the corresponding population parameter.
A. Standard error B. Sampling error C. Difference error D. None of the above |
ANSWER B
|
