HW #13 - O and E stuff (SOLUTIONS)
2. NO
If the digits were evenly distributed, the chance we would find evidence as strong or stronger than we got that they are not evenly distributed would be somewhere between 10% and 90% (closer to 90%). This is assuming the data was obtained in a proper fashion and all conditions were met.
3. Pips evenly distributed
Pips not evenly distributed
n = 11 + 7 + 9 + 15 + 12 + 6 = 60
each E is the same, namely np = 60(1/6) = 10
Pips / 1 / 2 / 3 / 4 / 5 / 6O’s / 11 / 7 / 9 / 15 / 12 / 6
E’s / 10 / 10 / 10 / 10 / 10 / 10
Picture of how would be distributed if the pips were evenly distributed. Note that the right hand tail is 10%. df = 6-1 =5.
NO
If the die was fair, the chance we would find evidence as strong or stronger than we got that it is not fair is somewhere between 10% and 90%. This is assuming the data was obtained in a proper fashion and all conditions were met.
7. NO
If the distribution of blood types were the same, the chance we would find evidence as strong or stronger than we got that they differ would be somewhere between 10% and 90% (closer to 90%). This is assuming the data was obtained in a proper fashion and all conditions were met.
8. Road kill rabbits 35% A, 25% B, 20% C and 20% D
Road kill rabbits not as above
n = 25 + 19 + 18 + 18 = 80
E’s are found by np = 80(.35) = 28, and 80(.25) = 20 and 80(.20) = 16
Rabbit Species / A / B / C / DO’s / 25 / 19 / 18 / 18
E’s / 28 / 20 / 16 / 16
Picture of how would be distributed if the pips were evenly distributed. Note that the right hand tail is 10%. df = 4-1 =3.
NO
If the road kill distribution was 35-25-20-20, the chance we would find evidence as strong or stronger than we got that it is notis somewhere between 10% and 90%(closer to 90%). This is assuming the data was obtained in a proper fashion and all conditions were met.
12. YES
If there is no relationship between age group and favorite fast-food hamburger restaurant, the chance we would find evidence as strong or stronger than we got that there is a relationship is less than .005. This is assuming the data was obtained in a proper fashion and all conditions were met.
13. How well flower seeds do and brand are independent
How well flower seeds do and brand are related
The bold face numbers are the O’s
Brand A / Brand B / Brand CDid not germinate / 15 / 20 / 9 / 44
Germinated, but no flowers / 29 / 30 / 19 / 78
Germinated, and flowered / 91 / 120 / 66 / 277
Totals / 135 / 170 / 94 / 399
The E’s are found by (row total)(column total)/(grand total) and are in bold face.
Brand A / Brand B / Brand CDid not germinate / (44)(135)/399 = 14.89 / (44)(170)/399 = 18.75 / (44)(94)/399 =10.37
Germinated, but no flowers / (78)(135)/399 = 26.39 / (78)(170)/399 = 33.23 / (78)(94)/399 =18.38
Germinated, and flowered / (277)(135)/399 = 93.72 / (277)(170)/399 =118.02 / (277)(94)/399 = 65.26
Picture of how would be distributed if they were independent. Note that the right hand tail is 1%. df = (3-1)(3-1) = 4.
=
=.986
NO
If the was no relationship, the chance we would find evidence as strong or stronger than we got that there is a relationship issomewhere between 90% and 95%. This is assuming the data was obtained in a proper fashion and all conditions were met.