Statistics in Excel Tutorial-1
Statistics in Excel Tutorial-1
Statistics in Excel Tutorial-1
0)
This workbook is designed to provide you a working model of how to perform statistics with
YOU MUST SAVE THIS TUTORIAL TO YOUR COMPUTER, THEN OPEN IT FROM YO
Please Maximize this Window to Full Screen size!
Table of Contents
Name A place to type your name
Install Directions for installing statistical analysis add-in
T1 Description of descriptive statistics
Rev Review of how to calculate means and standard deviations
T2 Description of inferential statistics
Hypo Explanation of Statistical Hypothesis Testing
Rej Explanation of Statistical Hypothesis Testing (cont'd)
Perf Instructions on how to perform Student's t-test
Tail Instructions on how to choose one-tail or two-tail t-test
Diff Instruction on how to determine significant difference
Report How to report results of Student's t-test in a lab report
Turn1 A practice problem.
Turn2 Another practice problem
Write How to write the results section of a lab report
Answers The answer sheet that you will print and turn in to your TA.
Please Note: If you are asked to click in a specific field (ie, "click in the YELLOW field"), please do a
clicked in another field, the tutorial will not work correctly!
YOU WILL BE REQUIRED TO COMPLETE THIS TUTORIAL AND PRINT AND TURN IN THE A
ndard deviations
ng (cont'd)
wo-tail t-test
difference
a lab report
rn in to your TA.
LLOW field"), please do as instructed. If you
work correctly!
NT AND TURN IN THE ANSWER SHEET
What's Your Name?
?QUESTION?
Your first question is "What's your name?" Type your first and
last name in the DARK GREEN field beside the word
"Answer:" below.
uestions to answer.
ur answers will be
SHEET. To write the
RK GREEN field
nswer. Keep it
4. Near the bottom of the Add-In window, click on "Go." Make sure
Excel Add-Ins is showing in the Manage drop-down menu.
5. Select both Analysis ToolPak and Analysis TookPak VBA in the
Add-Ins window and then click "OK."
6. After completing the above steps, click "Data" on the upper tool
ribbon and make sure you have Data Analysis box on the toolbar.
Congratulate yourself! You've installed the Statistics Add-In and are now ready to comp
MSExcel 2007)
Descriptive statistics are used to describe a single set of numbers or data. These statistics give
that you can use to describe a set of numbers or compare a set of numbers WITHOUT listing ev
The 2 main descriptive statistics we will use in this lab are…
b. Standard deviation - describes how spread out your numbers are around the mean on a num
if you subtract 2X standard deviation from the mean AND add 2X standar
mean, 95% of your data points will fall between those numbers. For exam
and the standard deviation is 5.1, 95% of all data points will be fall betwe
[2X5.1=10.2; 64.1-10.2=53.9; 64.1+10.2=74.3]
?QUESTION? ?QUESTION?
What is the function terms used to calculate If the mean is 100 and the standard dev
mean and standard deviation in MSExcel? between what two numbers will you find
all data points?
Answer: AVERAGE and STDEVA
Answer: 80 and 120
ology (Part 1)
r data. There are two main types of statistics
bs:
?QUESTION?
80 and 120
Review of Calculating Means and Standard Deviations in MSE
Please note: If you do not remember how to calculate means and standard deviations in MSEx
review the tutorial entitled "MSExcel Calculations and Routine Functions Tutorial."
Instructions:
Total Length (cm)Length 1st internode (cm)Root:Shoot Calculate the mean and
16.1 9.5 0.384 each column of data in th
15.2 11.2 0.369
Control 14.9 10.5 0.359 ?QUES
15.6 9.8 0.427
16.7 10.2 0.409 When writing a lab repor
Mean 15.7 10.24 0.3896 and standard deviation in
SD 0.71763500472 0.658027355054484 0.0281389 MEAN+/-SD UNITS. For
is 5.4 and standard devia
21.8 11.3 0.523 units are grams, you rep
25.7 9.6 0.489 "5.4+/-0.2 grams."
Auxin 20.5 9.8 0.589
23.2 10.5 0.455 How would you report th
22.4 12.4 0.429 total length of the control
Mean 22.72 10.72 0.497
SD 1.93571692145 1.15195486022674 0.062434 Answer:
Answer:
Deviations in MSExcel
dard deviations in MSExcel, please
ine Functions Tutorial."
Instructions:
?QUESTION?
15.7±0.72 cm
10.72±1.15 cm
Types of Statistics Used in Biology (Part 2)
2. Inferential Statistics
Inferential statistics are used to compare 2 or more sets of numbers or data. Two sets of numbe
due to random chance or can be actually different (for example, due to the effects of a chemical
Inferential statistics are used to determine whether or not 2 sets of numbers are different only du
actually different (=SIGNIFICANTLY DIFFERENT). There are a number of inferential statistical
various purposes. For this lab, we will use Student's t-test, which is used to compare 2 (and onl
?QUESTION? ?QUESTION?
In terms of the effects caused by the experimental Re-read the description of inferential
treatment, what does it mean when the experimental it mean when two sets of numbers (d
group is significantly different from the control? significantly different?
Answer: This means that the two sets of numbers are "a Answer: If the two sets of numbers
ology (Part 2)
?QUESTION?
If the two sets of numbers are not significantly different, the differences are due only to random chance.
andom chance.
Statistical Hypothesis Testing
When you perform inferential statistics (like the Student's t-test), you are calculating the probability that two sets
different due to random chance OR different due to real differences (for example, due to an effect caused by an ex
When two sets of data are statistically determined to be different due to real differences, we refer to the two sets
different. Therefore, we can formulate two different hypotheses (or two different possible reasons) that two se
1) That the differences are due to random chance only.
2) That the two sets of data are really different due to the effects of an experimental manipulation
data are significantly different).
The first hypothesis, that differences are due to random chance only, is called the null hypothesis (Ho). The secon
of data are significantly different, is called the alternate hypothesis (Ha). It is important to be able to formulate
hypotheses; if you are unable to do so it may mean that your data (or experiment) is not suitable for analysis b
In this tutorial, we will use some data collected from an experiment testing the effects of the addition of auxin on s
growth; for example, total length of the plant. If we were to compare total length of control and auxin-treated plant
not they are significantly different, we could develop Ho and Ha hypotheses. Our Ho hypothesis would be that any
lengths of control and auxin plants was due to random chance, that auxin did not have a significant effect on total p
controls. Our Ha hypothesis would be that the total lengths of control and auxin plants are really different, that au
effect on the total length as compared to controls.
?QUESTION?
You are performing an experiment where you are testing the effects of auxin on plant growth. You measu
internode of control plants and plants treated with auxin. You wish to compare the sets of data using
The lengthsFormulate
of the firstainternode
Ho and Hofa hypothesis.
control plantsWrite
and your
plantshypotheses
treated within auxin
the spaces provided
will differ only
Ho: The auxin will not have a significant effect on the length of the first internode of the auxin
The first internode lengths of the control and compared to the
auxin plants arecontrol plants.
actually different. The auxin did
Ha:
on the first internode length when compared to controls.
Testing
ting the probability that two sets of numbers or data are
ue to an effect caused by an experimental manipulation).
ences, we refer to the two sets of data as be significantly
nt possible reasons) that two sets of data are different:
So what does it mean for two sets of data to be SIGNIFICANTLY DIFFERENT? A significant difference is
random chance (Ho) is so low that we are reasonably confident that the sets of data are really different (H
situations, a 5% chance that the observed differences are random is sufficiently low to allow us to be rea
the sets of data are really different. Therefore, a p of 0.05 OR LESS is considered significantl
When there is a less than 5% chance that the differences are random, we can reject the Ho hypothes
reasonably confident that the Ha hypothesis is supported by our data. However, when there is greater tha
differences are random, we cannot be sufficiently confident and we fail to reject the Ho hyp
?QUESTION?
For the following p-values, determine whether or not you can reject the Ho hypothesis and whether or no
be significantly different.
p=0.23 p=0.0001 p=0.056
Reject Ho? Yes No Yes
Significantly different? No Yes No
Question…….
mulas. It is possible to perform the calculations
l to perform the Student's t-test. MSExcel will
umber of numbers, many of which you will not
ovide is the p-value (or p). The p stands for
ther words, p tells you the probability that any
nce. So, if you have a p of 0.34, there is a 34%
lly, there is a 66% chance that the two sets of
that the differences observed are random or a
eal.
p=0.056
Yes
No
Instructions for Performing the Student's t-test
Please Note: You MUST place the results of the Student's t-test in the YELLOW field in order for this t
Instructions for performing Studen
Total Length Length 1st internode Root:Shoot
16.1 9.5 0.384 1. Click on any field that has a NUMB
15.2 11.2 0.369 2. Click on "Data" on the toolbar nea
Control 14.9 10.5 0.359 3. Select the "Data Analysis" box fro
15.6 9.8 0.427 4. Select "t-test: two sample assumin
16.7 10.2 0.409 A box will appear entitled "t-test: T
Mean 15.7 10.24 0.3896 equal variances."
SD 0.71763500472 0.65802735505448 0.0281389 5. Click in the white box beside the w
6. Highlight the PINK fields which inc
21.8 11.3 0.523 7. Click in the white box beside the w
25.7 9.6 0.489 8. Highlight the BLUE fields which in
Auxin 20.5 9.8 0.589 experimental group.
23.2 10.5 0.455 9. Click on the button beside the wor
22.4 12.4 0.429 10. Click in the white box beside the w
Mean 22.72 10.72 0.497 11. Select any field that does not con
SD 1.93571692145 1.15195486022674 0.062434 For example, select the YELLOW
12. Check to make sure you entered
t-Test: Two-Sample Assuming Equal Variances 13. Click on "OK."
14. Go to next page!
Variable 1 Variable 2
Mean 15.7 22.72
Variance 0.515 3.747
Observatio 5 5
Pooled Var 2.131
Hypothesiz 0
df 8
t Stat -7.6035322375
P(T<=t) one3.1420631E-05
t Critical o 1.85954803753
P(T<=t) two6.2841262E-05
t Critical tw 2.3060041352
Student's t-test
LOW field in order for this tutorial to work correctly!
ns for performing Student's t-test:
next page!
One-tail or Two?
Please Note: The results printout below will have all 0's if you did not place the results of the t-test in the YELLOW field as ins
Before you can determine whether or not two sets
t-Test: Two-Sample Assuming Equal Variances different, you need to know whether or not you ne
Total Length: Control vs. Auxin t-test. Your hypothesis that you develop for your e
Variable 1 Variable 2 which is appropriate. A one-tail t-test is used whe
Mean 15.7 22.72 treatment will cause the mean of your experimenta
the mean of your control group OR that the mea
Variance 0.515 3.747
group will be lower than the means of the control
Observations 5 5 your hypothesis is that if you treat plants with a h
Pooled Variance 2.131 larger than untreated plants, you are hypothesizin
Hypothesized Mean 0 experimental group will be higher than the mean o
df 8 you need to use a one-tail t-t
t Stat -7.60353
P(T<=t) one-tail 3.14E-05 A two-tail t-test is used when you hypothesize
t Critical one-tail 1.859548 experimental group will be different (higher or low
P(T<=t) two-tail 6.28E-05 control group. When using a two-tail t-test, you
t Critical two-tail 2.306004 notion of whether the mean will be higher or low
different. For example, if your hypothesis is that i
hormone the size of the plant will be different tha
plants, you need to use a two-ta
ANOTHER NOTE: Notice how this results When you perform the Student's t-test with MSEx
printout is labelled (yellow fields) so that you with the results of both the one-tail and two-ta
know what was compared. determine which you need to use and refer to the
THIS DOES NOT HAPPEN AUTOMATICALLY! results print-out.
You will assume that you need to use one-tail t
of this tutorial.
?
e t-test in the YELLOW field as instructed on the previous page!
mine whether or not two sets of data are significantly
know whether or not you need a one-tail or two-tail
is that you develop for your experiment will determine
A one-tail t-test is used when you hypothesize that a
he mean of your experimental group to be higher than
ontrol group OR that the means of the experimental
han the means of the control group. For example, if
hat if you treat plants with a hormone they will grow
d plants, you are hypothesizing that the mean of your
will be higher than the mean of the control. Therefore,
you need to use a one-tail t-test.
?QUESTION?
0.220944
-0.80904
No
It's YOUR Turn! (Part 2)
Your Assignment:
Total Length Length 1st internode Root:Shoot Perform the Student's t-tes
16.1 9.5 0.384 ratio of the auxin treated g
15.2 11.2 0.369 You must place the printou
Control 14.9 10.5 0.359 field.
15.6 9.8 0.427
16.7 10.2 0.409
Mean 15.7 10.24 0.3896 ?QUES
SD 0.717635 0.65802735505448 0.02813894
From the results printout, w
21.8 11.3 0.523
25.7 9.6 0.489 Answer:
Auxin 20.5 9.8 0.589
23.2 10.5 0.455 What was the t Stat?
22.4 12.4 0.429
Mean 22.72 10.72 0.497 Answer:
SD 1.93571692 1.15195486022674 0.06243397
Are the two sets of data sig
t-Test: Two-Sample Assuming Equal Variances
Answer:
Variable 1 Variable 2
Mean 0.3896 0.497
Variance 0.0007918 0.003898
Observations 5 5
Pooled Variance 0.0023449
Hypothesized Mean 0
df 8
t Stat -3.5068091
P(T<=t) one-tail 0.00399944
t Critical one-tail 1.85954804
P(T<=t) two-tail 0.00799888
t Critical two-tail 2.30600414
t 2)
Your Assignment:
?QUESTION?
0.003999
-3.50681
Yes
Reporting Your Results! Putting Means, Standard Deviations, and Stat
When you write the results section in your lab report, you will be required to present the means, standard
of statistical analyses. Please refer to the page entitled "Rev" for information on how to present means an
and the page entitled "Report" for information on how to present the results of the Student's t-test.
The best (easiest to read and most acceptable for a good grade) is to combine the means, standard devia
of the Student's t-test into a sentence together. The following example is of the data for total length of pla
the auxin treated group.
"The mean total length of the control plants was 15.7+/-0.72 cm. The mean total length of the
auxin was 22.7+/-1.94 cm, which was significantly different from the control (p=3.14E-0
Please note: Even if the sets of data are NOT SIGNIFICANTLY DIFFERENT, you STILL report the result
?QUESTION?
Write a sentence or two that reports the means, standard deviations, and results of the Student's t-test for
You will need to go back to previous pages to obtain the numbers. Don't forget to include the results for c
Length of the 1st internode is measured in centimeters (cm)!
Answer: The mean 1st internode length of the control plants was 10.24± 0.65 cm . The mean 1st interno
Write a sentence or two that reports the means, standard deviations, and results of the Student's t-test for
Root:shoot (root to shoot ratio) has NO units!
Answer: The mean 1st internode length of the control plants was 0.390 ± 0.03. The mean 1st internode
Deviations, and Statistics Together
present the means, standard deviations, and the results
on how to present means and standard deviations
f the Student's t-test.
ne the means, standard deviations, and the results
he data for total length of plants in the control and
65 cm . The mean 1st internode length of the plants treated with auxin was 10.72 ± 1.15 cm , which was not significan
Instructions: Print this page and turn it in to your Instructor! Do not type anything on
this page.
If the two sets of numbers are not significantly different, the differences are due only to
T2-2 random chance.
The lengths of the first internode of control plants and plants treated with auxin will differ
Hypo-1 only due to random chance. The auxin will not have a significant effect on the length of the
first internode of the auxin treated plants when compared to the control plants.
The first internode lengths of the control and auxin plants are actually different. The auxin
Hypo-2 did have a significant effect on the first internode length when compared to controls.
The mean 1st internode length of the control plants was 10.24± 0.65 cm . The mean 1st internode length of the
Write-1 plants treated with auxin was 10.72 ± 1.15 cm , which was not significantly different from the control (p=0.22 ,t=
-0.81).
The mean 1st internode length of the control plants was 0.390 ± 0.03. The mean 1st internode length of the plants
Write-2 treated with auxin was 0.50 ± 0.06 , which was not significantly different from the control (p=0.004 ,t= -3.51).
Tutorial Completion #:
0.658027 0.497 1.859548 0.88 1.859548