Rehabilitation Programs Assignment

This second assignment is used to prepare students on kind of questions to

expect in the final professional examination

For the first 15 questions use instructions below

- Answer all questions
- For each statement in the questions use T for True and F for False
- Correct answer 1 mark: Wrong answer -1/2 marks; unmarked 0 marks
- This section carries 15 marks, marks will be deducted for wrong answers
as above

Example. In hypothesis testing for a single population, we use the Z test if: -
T F (a) the sample size is small
T F (b) the population standard deviation is known
T F (c) the sample size is large
T F (d) we do not know the degrees of freedom
T F (e) we know the level of significance

How you are supposed to respond on your answer sheet

Example answer (a) F (b) T (c) T (d) F (e) T

Answer in this manner all the following 15 questions [Answers should fit on one page or
less]

1. The following are properties of a normal distribution

T F (a) Mean = Median(M) = Mode(Z)

T F (b) It can have two modes thus become bi modal
T F (c) Some portion of curve lies below x-axis
T F (d) Not all theories of sample tests t, f and chi-square test are
based on the normal distribution
T F (e) The mean of the distribution does not determines the
location of the centre of the graph

2. The purpose of hypothesis testing is: -

T F (a) to reject the null hypothesis

T F (b) to test for the significance of a sample
T F (c) to test for the significance of a population parameter
T F (d) to measure the difference between samples
T F (e) to reach a conclusion about a population parameter using a
sample

3. You wish to test the null hypothesis that two population means are equal for a
certain variable. Which ONE of the following conditions must apply before one can
justifiably carry out a t-test in order to test this null hypothesis

T F (a) the variable must have a normal (approximately) distribution

T F (b) at least one of the same sizes must be equal to or larger
than 30
T F (c) there should be equal sample sizes in the two samples
T F (d) the data must be ordinal
T F (e) the data from the two samples must be paired

4. The 95% confidence interval for hematocrit in men is 42.2 to 49.2

 T F (a) any man with hematocrit outside the limits is abnormal
T F (b) hematocrits outside these limits are proof of disease
T F (c) a man with a hematocrit of 16 must be very healthy
T F (d) we are 95% sure that hematocrit of man pick from this same
population will fall between 42.2 and 49.2
T F (e) a man with a hematocrit of 42 may be ill

5. The one sample t-test is appropriate when

T F (a) our aim is to compare the mean of a variable in one group of

individuals to that of another
T F (b) our aim is to compare the mean of a variable in a group of
individuals to a particular value
T F (c) the variable of interest is categorical
T F (d) the variable of interest in not normally distributed
T F (e) the data is skewed

6. A random sample from an unknown population had a standard deviation of zero.

Which of the following is a reasonable conclusion

T F (a) the sample range is zero

T F (b) a mathematical error was made in computing the sample
standard deviation. It must be greater than zero
T F (c) the populations standard deviation must be large
T F (d) the popular standard deviation must be small
T F (e) the sample range is infinitely large

7. Say whether the following statements are true or false

T F (a) The level of significance used in statistical significance is

always p=0.05
T F (b) p-value of 0.001 shows that differences being assess are real
T F (c) p-value does not always help to reject the null hypothesis at
a set level of significance
T F (d) the lower the p-value the more likely results will benefit the
patients
T F (e) p value for a relative risk does not relate to confidence
interval of a relative risk

8. In the presentation of data

T F (a) A pi chart is the one in which a circular ‘pie” is split into

sectors, one for each category of a categorical variable, so
that the data of each sector is equal
T F (b) We use the median and interquartile range only when the
data is normally distributed
T F (c) The distribution of a variable is left skewed if a histogram of
observed values has a long tail to the right with one or few
high values
T F (d) The ends of the whiskers mark the interquartile range
T F (e) Age can be presented in 5 year age ranges but they are not
ordinal

9. The following can be said about anova

 T F (a) It is used to compare a categorical or continuous variable
across more than 2 groups
T F (b) The degrees of freedom do not matter
T F (c) If there are three groups the F statistics can be significant
even if only one set of the groups in different
T F (d) ANOVA is not an extension of a test
T F (e) Kruskal Wallis Test can be used as a replacement of ANOVA
in all cases

10. The following can be said about parametric and non-parametric tests

T F (a) They use the same assumptions to for them to be valid

T F (b) They replace each other in all cases
T F (c) Parametric tests have statistical rigor compared to no
parametric test
T F (d) Nonparametric test makes analysis of results easier
T F (e) A repeated measures ANOVA can not be used is used when
the same participants are measured multiple times under
different conditions

11. The following can be said about relative risk

T F (a) It can not be used in follow up studies

T F (b) If the health conditions is lower in the exposed groups the
statistic seems to be less valid
T F (c) The relative risk can never be one
T F (d) You can not use relative risk together with p values
T F (e) It can show is the exposure is results in an increase or
decrease in the outcome of interest

12. In a chi square test the determine the association between gender and HIV the
following can be said

T F (a) H0: Gender and HIV Status are not independent.

T F (b) H1: Gender and HIV Status are independent
T F (c) The chi-square formular is X2 = Σ(O-E-0.5)2 / E
T F (d) Odds ratio of 2.2 for HIV prevalence among females
compared to males says HIV is higher among females than males
if the p value is 0.061
T F (e) If the p-value that corresponds to the test statistic X 2 with
(#rows-1)*(#columns-1) degrees of freedom is
less than your chosen significance level then you can reject
the null hypothesis.

13. A t-test was used to compare the association between number of blood
transfusion and death at 1 year among children with aplastic anemia. The
following results where obtained

. ttest bloodtransfusion, by( dead_alive_1_year )

Two-sample t test with equal variances

Group Obs Mean Std. err. Std. dev. [95% conf. interval]

0 26 12.07692 1.32772 6.77007 9.342433 14.81141

1 23 5.217391 .8527021 4.089416 3.448995 6.985787

Combined 49 8.857143 .9422076 6.595453 6.962708 10.75158

diff 6.859532 1.624527 3.591405 10.12766

 diff = mean(0) - mean(1) t = 4.2225
H0: diff = 0 Degrees of freedom = 47

Ha: diff < 0 Ha: diff != 0 Ha: diff > 0

Pr(T < t) = 0.9999 Pr(T > t) = 0.0001 Pr(T > t) = 0.0001

T F (a) The mean number of blood transfusion is higher among

those who are alive compared to t hose who are dead
T F (b) The confidence intervals for the number of blood transfusion
overlap
T F (c) The p value is very low and its usually unexpected
T F (d) The groups were supposed to have equal numbers for the
test to be valid
T F (e) The standard error is a useless statistic in this scenario

14. The following can be said about probability

T F (a) The probability of an event is a number between 0 and 1

that expresses how likely it is to occur
T F (b) Empirical Probability ( or experimental probability) is based
on a mathematical model of probability
T F (c) Therefore, the range of all possibilities is 0 ≤ P ( A ) ≥ 1
T F (d) It is not certain that you will roll a number between 1 and 6
if you roll a six-sided die
T F (e) If two events, A and B, are mutually exclusive, then the
following formula can be used P ( A ∪ B )=P ( A )−P(B).

15. The following can be said about relative risk

T F (a) It can not be used in follow up studies

T F (b) If the health conditions is lower in the exposed groups the
statistic seems to be less valid
T F (c) The relative risk can never be one
T F (d) You can not use relative risk together with p values
T F (e) It can show is the exposure is results in an increase or
decrease in the outcome of interest

Answer the following questions

1. A study was conducted to investigate the effectiveness of bicycle safety helmets in

preventing head injury. The data consist of a random sample of 793 persons who were
involved in bicycle accidents during a one‐year period {Sample from Past Test]

Wearing Head injury Total

helmet
Yes No
 Yes 17 130
No 218 428
Total

(a) Calculate the marginal totals to the above Table. [2]

(b) Calculate the proportion of cyclers with head injurie in each group. [2]
(c) Test the null hypothesis that there is no difference in the proportion of cyclers with
head injuries in both groups.
[10]

2. Kindergarten students were the participants in a study to determine the fine motor
skills of 37 children receiving occupational therapy. An index of fine motor skills that
measured hand use, eye–hand coordination, and manual dexterity before and after 7
months of occupational therapy was used. Higher values indicate stronger fine motor
skills. The scores appear in the following table.

Subjec
t Pre Post
1 91 94
2 61 94
3 85 103
4 88 112
5 94 91
6 112 112
7 109 112
8 79 97
9 109 100
10 115 106
11 46 46
12 45 41
13 106 112
14 112 112
15 91 94

16 115 112
 17 59 94
18 85 109
19 112 112
20 76 112
21 79 91
22 97 100
23 109 112
24 70 70

(a) Giving reasons state whether these are independent or dependent samples [2].
(b) State the null and the alternative hypothesis to test.
[2]
(c) State the test statistic and the decision rule for the test.
[2]
(d) State the assumptions for the valid application of the test statistic in (c) above.
[4]

The following data represent diastolic blood pressures taken during rest. The x
values represent the length of time in minutes since rest began, and the
y values denote diastolic blood pressures (dbp).
x (time) 0 5 10 15 20

y (dbp) 72 66 70 64 66

(a) Describe four assumptions that do not need to be violated to calculate a

regression model so show correlation between these values
[4]

(b) If the relationship has a R2 value of 0.75 explain what this means [1]

Last question: Compare and contrast any two non- parametric tests and their
parametric equivalent [6]