MIDLANDS STATE UNIVERSITY
FACULTY OF SCIENCE AND TECHNOLOGY
APPLIED BIOSCIENCES AND BIOTECHNOLOGY
DEPARTMENT
ADVANCED STATISTICS FOR BIOLOGISTS
CODE: HABB 401
SESSIONAL EXAMINATIONS |
OCTOBER-NOVEMBER 2019 |
DURATION: 3 HOURS |
EXAMINER: Dr T. Muteveri |
SSeS |
INSTRUCTIONS |
1. Answer Question No. (SECTION A) and any
THREE questions from SECTION B.
2, Bach question carries 25 marks. Where a question
contains subdivisions, the mark value for each
‘subdivision is given in brackets. Illustrate your
nswer, where possible, with large clearly labelled
diagrams.
3, This question paper consists of 12 printed pages |
including appendices
T[/WABB 401~ SECTION A
In an effort to improve their programme for teaching stud
instructors conducted the following experiment. T hey rand
who registered to take first-semester French and randomly,
students each. Group 1 is taught French the trad
‘a programmed text, Group 3 is taug!
French from films of people from Franc
‘Table 1 French scores
Sroup1 Group2 Group} Group 4
5 80 87
10 B 65 90
90 70 70 85
80 60 68 18
78 65 R 80
{a) Identify the respons
(b) Test the data for con!
{@) Explain whether the factors in this study Follow fixed
(g) Perform an appropriate test to determine if teaching met
French.
(¢) Report the findings of this study.
and predictor va
ormity to normality and homogen
{@) Why is it important to determine sample size as part of
(b) Discuss factors that influence sample size.
{c) How would you es
imate sample size for a study
3, (a) Briefly discuss the rationale and objective:
{b) Describe how you would use the followin
data transformation:
(i) ladder of powers,
(ii) Taylor's power law, and
Gi) experience.
jents how t
ional lecture method,
ht French from tape-recorded |
.e. At the end of semester,
oral final examination on ability to speak French. The scores in Table
10 speak French, MSU language
jomly selected 20 students from all those
divided them into for groups of five
Group 2 is taught French from
lessons, and Group 4 is taught
all 20 students are given the same
1 were recorded.
riables in this experiment. (21
icity of variances. Is}
fects or random effects model. [3]
shod has an effect of ability to speak
{10}
15]
SECTION B
‘experimental design? {61
(13)
‘comparing means of two groups. {6}
sof data transformation {10}
1 approaches in choosing an appropriate
(5]
(3)
15]
— a PIWABB 401Six greenhouse benches were set up as blocks. Within each block, on® of each of four varieties of house
plants was planted. The plant heights (in centimetres) attained are tabulated in Table 4. Study the data
caarable 4 and the SPSS output in Appendix 4 before you answer the questions below
‘Table 4 Plant heights (cm) attained by different varieties
a Variety __
Block tS
1 19.8 21.9 16.4
2 16.7 19.8 154
3 177 210 14.8
4 18.2 214 15.6
5 20.3 22.1 16.4
_ 6 155 20.8
(a) State the factor(s) and the response variable(s) in this study. 4]
{b) Identity the experimental design used in this study. Explain your answer GB)
{¢)_ What does the means plot suggest about the effect of variety on plant height? BI
{@) Conduct the necessary test to determine if variety affects the plant height and perform
multiple comparisons if necessary. [15]
Write short notes of the following:
(a) stratified random sampling. 7
(b) interaction of factors, 15]
(c) role of statistics in the scientific method of research, and {8}
(d) the p-value approach in hypothesis testing. 15)
S|HABB 401research paediatrician is interested in predicting the birth weight of first-born infants from the
arcunt of weight gained by the mother during her pregnancy for the population of middle-class
arfviduals, She therefore randomly selects 15 middle-class mothers, pregnant for the first time,
inet follows them through pregnancy. After the mothers give birth, she records their weight gains
im pouinds and the birth weights of their offspring in pounds. The data are given in Table 6, Use
the data in Table 6 and the SPSS output in Appendix 6 to answer the questions below.
‘Table 6 Relationship between weight gain of mother and birth weight of infants
13 3.2
10.8 28
153 3.
95 27
13.5 3.6
14.4 3.7
17 34
12.2 34
0 3A
(a) Comment on the tenability of linearity, normality and homogeneity of variances to this data
sel. 61
(b) Construct the linear regression equation for predicting the birth weight of an infant from ane )
weight gain of the mother. 3]
(c) Test for the significance of the regression coefficients of the equation in (b). 6)
(d)_ Test for the usefulness of the regression equation in (b). BI
(e) Interpret the slope of the regression equation in (b), PI
(1) G)_ Predict the birth weight of an infant whose mother gained 12.0 kg 2]
(ii) Why would it not be advisable to use your regression equation to prediet the birth weights of
infants whose mothers’ weights are below 5.4 kg and above 15.3 ke? (31
4,HABB 401°APPENDICES
‘Appendix 1: SPSS output for Questi
Test of Homogeneity of Variances
reaching method Kolmogorov: Sminov* Shapio-Wik
tea atisic | of _| Sig] _Stotistc_| sig._|
“Traattonal methos 254 5| 200 914 492
score | Poafanmedion ee ee 942
| Tape-recoreded essons | _230|__——s|_—200|__934 625
Films 206|5|_—_—-200 963) ar
+ This is a lower bound of the true significance.
2 Liliefors Significance Correction
‘ANOVA
781.800
598,000
1379.800
S|HABB 401Deseriptives
Teaching method
Statistic_| Std. Error
Score
‘Traditional method
Programmed text
‘Mean
7e.00| 3.381
‘95% Confidence Interval for |Lower Bound _|
| Upper Bound
Mean
| 5% Trimmed Mean.
Median
| Variance
[Sid Deviation
Minimum
Maximum
Range
|interquartie Range __
Mean
95% Confidence Interval for [Lower Bound
Mean _ Upper Bound
5% Trimmed Mean
Medien _
Variance
‘Std. Deviation
Minin
im
Maximum _
Range.
195% Confidence Interval for
fm «Bound _|
Mean Upper Bound
5% Trimmed Mean
Tape-recoreded lessons Neston =
| variance
‘to. Deviation
Minimum [ 65
[Maxum
i Range
Mean —
95% Confidence Interval for ju sound | 76.02)
Mean Uppersouna | 9078
rime ‘s% Trimmed Mean 2 350]
| Mecian 25.00
Vanance
| std Deviation _
Minimum 1s _|
6|/HABB 401| Maximum
Range
4s]
T/HABB 401Plant height (em)
Type Ill Sum of
Squares
‘é 7024 682
19.793
198538
6527
7239.540|
214.858
‘a. R Squared = 970 (Adjusted R Squared = 953)
Variety
Variety +
(—varelty2
| == Variety 3
Variety 4
ABB 401a Mattie Comparisons
[Dependent Varianie height pr
Hrukey HSD
p Variety [() Variety | Mean | Std. Error
ference hs) |
2 arog] 3008
1 3 2.500" 3808
E 4 4300] 3808)
1 3.133'| 3608!
2 3 se3s| 9808
4 167
1 “2.500
3 2 5635)
ae 1 =
4 4.333, 3808 |
4 2 24er| 3008
| a _-1893'|__.3808|_
Based on observed means
“The error fem is Mean Squares) = 496
[+ Tne mean aferenceis significant atthe 005 level, —__|
9/WABB 401Kolmogorov-Smitnov®
‘Statistic Sig
Shapico-Wike
ot
* This is a lower bound of the true significance
1a Variety = Variety 2
+ This is @ lower bound of the true significance.
a Variety = Variety 3
‘Tests of Normality
+ This is a lower bound of the true significance
1. Variety = Variety 4
Statistic
TO/HABB 40“Appendix 6: SPSS Output for Question 6
a
birth weight of infant
30
Unstandardized Residual
50 7 0 125 160
weight gain of mother
eee
yedhoo sadso0 aon —«30on0 = (seNo00
Unstandardized Predicted Value
3acho
Tij/WaBB 401‘3 Depenident Variable: birth weight of infant
b Allrequested variabies entered
Model Summa
Adjusted R
Square
R Square
‘2. Predictors: (Constant), weight gain of mother
». Dependent Variable: birth weight of infant
Regression
Residual
Total
‘a Dependent Variable: bith weight of infant
bb Predictors: (Constant), weight gain of mother
(Constant)
‘a, Dependent Variable: birth weight of infant
‘Std. Error of the
Esimate,
Coefficients?
Unstandardized Coefficients
1648
[| ee
Standaraized
Costficients
END OF QUESTION PAPER
~12;HABB 401
J
Origin of Life Theory Source: The Science News-Letter, Apr. 29, 1961, Vol. 79, No. 17 (Apr. 29, 1961), P. 263 Published By: Society For Science & The Public