Quiz
Hypothesis testing
This quiz has question types that open in the usual player.
Questions in this quiz (28)
1What is meant by a type I error?
- Arejecting a true null hypothesis
- Brejecting a false null hypothesis
- Cfailing to reject a false null hypothesis
- Dfailing to reject a true null hypothesis
2What is meant by a type II error?
- Arejecting a true null hypothesis
- Bfailing to reject a true null hypothesis
- Crejecting a false null hypothesis
- Dfailing to reject a false null hypothesis
3In a one sample t test, what would the null hypothesis be?
- Athe mean of the sample is equal to the hypothesised value
- Bthe mean of the sample is not equal to the hypothesised value
- Cthe means of this set of differences is equal to zero
- Dthe means of this set of differences is not equal to zero
4In a paired t test, what would the alternate hypothesis be?
- Athe mean of this set of differences is equal to zero
- Bthe mean of this set of differences is not equal to zero
- Cthe mean of this set of differences is equal to the hypothesised value
- Dthe mean of this set of differences is not equal to the hypothesised value
5For a pair of measurements per subject (e.g. before and after values), it would be best to use a [blank_start]paired t test[blank_end].
- Apaired t test
- Bone sample t test
- Ctwo sample t test
- DANOVA
6A paired t test works by converting pairs of values into differences, and then tests whether the differences are equal to zero, using a one sample t test.
- ATrue
- BFalse
7In a two-sample t test, the null hypothesis would be: the [blank_start]means[blank_end] of these two samples are [blank_start]the same[blank_end].
- Ameans
- Bmean of this set of differences
- Cthe same
- Ddifferent
- Eequal to zero
8To use a two sample t test there must be:
- Aequal variances
- Bno variance
- Cat least 30 samples per group
- Dless than 30 samples per group
9If there is not equal variances, what test could you use instead of a two-sample t test?
- AANOVA
- BMann-Wilcoxon-Whitney
- CPaired t test
- DWelch's t test
10In an ANOVA test, what would the alternate hypothesis be?
- Aall of the means are different to each other
- Bat least one of these groups has a different mean from the others
- Call of the means of the groups are the same
11What is the test statistic of an ANOVA?
- AF
- BP
- Cq
12What does the test statistic of ANOVA mean?
- AF = treatment mean square / error mean square
- BF = error mean square / treatment mean square
13Treatment mean square = explained variation = 'signal'
- ATrue
- BFalse
14Error mean square = [blank_start]unexplained[blank_end] variation = 'noise'
- Aunexplained
15What is meant by the total sum of squares?
- Asquared deviations from the grand mean
- Bsquared deviations from the error
- Ctreatment sum of squares + error sum of squares
16Treatment sum of squares = squared deviations from the grand means to the [blank_start]group means[blank_end]
- Agroup means
17Residual / error sum of squares = squared deviations from group means to [blank_start]data points[blank_end]
- Adata points
18How would you calculate total degrees of freedom?
- Atreatment d.f. + error d.f.
- Btreatment d.f. + sum of squares
- Ctreatment d.f. - error d.f.
19ANOVAs can be used to compare means of two or more groups, and interactions.
- ATrue
- BFalse
20X2 tests are made to look for
- Ainteractions
- Bdifferences between pairs
- Coutliers
21Which of the following are assumptions of a chi-squared test?
- Aexpected counts are >5
- Bexpected counts are <5
- Cmeans of counts = variance of counts
22If you test for multiple hypotheses on the same dataset, you must correct for multiple comparisons.
- ATrue
- BFalse
23To carry out a Bonferroni correction, if the overall significance value is [blank_start]a[blank_end] and you carry out [blank_start]m[blank_end] hypothesis tests, the p value must be < [blank_start]a/m[blank_end]
- Aa
- Bm
- Ca/m
24What is the main assumption of a t test?
- Aassumes data is normally distributed
- Bassumes data is not normally distributed
25What are the main assumptions of an ANOVA?
- Aassumes unexplained variation is normally distributed
- Bassumes variance of unexplained variation is constant throughout the dataset
- Cassumes variance of explained variation is constant throughout the dataset
- Dassumes explained variation is normally distributed
26If the data doesn't fit the assumptions for an ANOVA, you can:
- Atransform the data, (e.g. by square rooting)
- Buse a test that makes fewer assumptions, (e.g. Mann-Whitney-Wilcoxon)
27What is the non-parametric equivalent of the two sample t test?
- AWelch's t test
- BMann-Whitney-Wilcoxon test
- CRandom Forest regression
- Dlinear regression
28What is the null hypothesis of the Mann-Whitney-Wilcoxon test?
- Athe two groups being compared come from the same distribution, with the same median
- Bthe two groups being compared come from a different distribution, with a different median
- Cthe two groups have a mean of zero
- Dthe two groups have a mean that is not equal to zero
Information
- Description
- MIBTP Statistics Quiz on Hypothesis testing, created by Bethany Richmond on 04/11/2018.
- Author
- Bethany Richmond
- Language
- English
- Views
- 0
- Tags
- Rating
- No ratings yet Rate this quiz