Question

Stath Question..

Answer 1

Part a and b are mostly just vocab, in general what does ''discrete'' mean?

Answer 2

a. Do you know what is meant by discrete vs. continuous variable? Basically, if it can be "counted" then it is generally discrete. Continuous is something that is "measured" on a continuous scale, and limited only by the accuracy of the instruments.

b. What do you think? Have you looked up the definition of each type of sampling?

c. Here you are just to discuss some of the shortfalls of mailed questionnaires. Those are probably discussed in this section of your course materials, but you can probably think of some obvious ones. Low response rate and cost are obvious ones.

Answer 3

@cinar in a way, specifically any whole numbers or ''counting'' numbers and no fractions or decimals continuous means there are no breaks so you could have any type of numbers (whole numbers or decimals)
So you do think# of shares only uses whole numbers or can one person hold say.4 of a share?

Answer 4

As I mentioned before, discrete means whole numbers whole continuous means we can have whole numbers or decimals/fractions. So if you are thinking about stocks, they sell individually. A person can't only like 40% of one stock
* A person can't own 40% of one stock
So since you can't have fractions of stocks, do you think shares are discrete or continuous?

Answer 5

a) is discrete

Answer 6

Yep! Because you can't have fractions of items, good job!

Answer 7

thanks (:

Answer 8

Any guess on b, or you want me to explain it?

Answer 9

let me see

Answer 10

Sure

Answer 11

it seems random, but we have not learned other definition though

Answer 12

Have a guess at each term ;) What do you think simple random sampling is?

Answer 13

What does ''random'' mean in math?

Answer 14

every one has equal chance

Answer 15

Exactly!

Answer 16

Simple random sampling means every single thing has an equal chance of being chosen. Do you think that is what it is if we pick 20 people from 50 states?

Answer 17

"If the survey is conducted by telephoning 20 randomly selected " this implies random sampling right

Answer 18

Every single type of sampling there involves random selection at some point but not random sampling.
By picking only 20 people from every state you are already limiting the ''equally likely'' part. You are influencing the outcome by limiting x people from every state. So its not random sampling because every person is not equally likely

Answer 19

I have no idea about this term "stratified"

Answer 20

Stratified means that you break up your population according to some criteria and then pick a random sample from each resulting "strata".

Answer 21

Stratifying refers to breaking things into subgroups

Answer 22

Like, if I wanted to sample students on a college campus, and I randomly choose 10 from each dorm. That is stratified.

Answer 23

then what is the criteria for this question.

Answer 24

Any thoughts on systematic or convenience sampling?

Answer 25

I know systematic sampling

Answer 26

convenience sampling?

Answer 27

You want to go through each definition an pick what fits best.
What type of sampling you have depends on the context of the question, so what do you think systematic sampling is based on the name?

Answer 28

that expling everthing @DebbieG

Answer 29

also enlighten me with what systematic sampling is in your view?

Answer 30

:O

Answer 31

well you are choosing your sample in some order

Answer 32

say choose 10 student in biology and then 10 student in math son son

Answer 33

Right, ''order'' is the key term there systematic usually involves a list

Answer 34

I get you :)
But did they make a list here?

Answer 35

"randomly selected stockholders "

stockholders is our dorm (:

Answer 36

no that is not in case

Answer 37

Yep! So not random sampling or systematic
From the name, what do you think ''convenience sampling'' is?

Answer 38

stratified random sampling is the answer

Answer 39

Why?

Answer 40

let me understand first meaning of convenience sampling

Answer 41

You want to have all the definitions in front of you first, just in case there are some with overlapping criteria

Answer 42

stratified random sampling, because your smaple contains only stockholders

Answer 43

That's not what makes stratified random sampling different from the other definitions though

Answer 44

I believe convenience is not random at all

Answer 45

stratified random sampling would need many shared criteria between the individuals in the population

Answer 46

That's right...it's in the name ''convenience''...they would just pull anyone there. But what about ''cluster''? What does that mean?

Answer 47

? @cinar

Answer 48

sorry..

Answer 49

no idea :(

Answer 50

The general usage of cluster can mean ''group of similar objects growing closely together'' but in math, cluster refers to a group
The definition of cluster sampling specifically mentions grouping participants by geography or time
That's the one we want
stratified cluster sampling are closely related but stratified sampling has more similar features between participants
cluster sampling specifically involves time or geography
Does it make sense why part b is cluster sampling?

Answer 51

himm

Answer 52

do you not understand, is it the matter?

Answer 53

kinda

Answer 54

What part are you confused at?

Answer 55

then key point is difference between stratified and cluster sampling

Answer 56

but stratified sampling has more similar features between participants? such as?

Answer 57

here, yes

Answer 58

like grouping by age

Answer 59

or grouping by race

Answer 60

I see

Answer 61

Whenever you have a question like this you want to run through the definitions before drawing a conclusion s you identify the ''best'' choice
Ready for c?

Answer 62

what is similar features of stockholders ?

Answer 63

you can use age or race or gender
but it would be like multi-layers
that's what you call ''strata''

Answer 64

sorry I m international student..

Answer 65

you could also probably stratify by income too since its stockholders

Answer 66

ok

Answer 67

ready for c

Answer 68

I'll be back in 15 min sorry..

Answer 69

I think the sampling here is hard to characterize. I see it as stratified, with the stratas being each state. That is how the the population is being broken up.

I don't think it's cluster, because in cluster sampling, you sample EVERY element in the chosen clusters. Like, if I made the clusters each floor of each dorm on campus, and then I randomly choose 3 floors and sample everyone on those floors.

Although the strata of "states" isn't as clearly "stratified" as some other options might be, I do think that stratified is the best answer.

Answer 70

My understanding of ''cluster'' is it has to do with ''natural'' groups.
In my stats text, cluster specifically mentions geography & time so that's why I think that is the best choice. When i think of stratified, I think of multiple subgroups.
Like they would have to break the stockholders up with male vs female, and by age together

Answer 71

@DebbieG

Answer 72

like here: http://www.princeton.edu/~achaney/tmve/wiki100k/docs/Cluster_sampling.html

Answer 73

The important part is the ''total pop is divided into these groups and a sample of groups is selected''
That's why I think cluster is the best choice, it seems like they pick the states first then pick the people.

Answer 74

I agree it isn't the "typical" stratified situation, but isn't the typical cluster sampling either.

As your link states: "The main difference between cluster sampling and stratified sampling is that in cluster sampling \( \large \color{red}{\text{the cluster is treated as the sampling unit}}\) so analysis is done on a population of clusters (at least in the first stage)."

Everything in the chosen clusters is sampled in cluster sampling. Like you said "'total pop is divided into these groups and \( \large \color{red}{\text{a sample of groups is selected''}}\)

That isn't what is happening here, there is a random sample taken within each state. That sounds more like stratified to me.

But again, I don't think it's a perfect example of either, stratified or cluster.

The book I am currently teaching stats from has these definitions:

Stratified sampling: divide the population into groups (caled strata) according to some characteristic that is important to the study, then sample from each group. SAMPLES WITHIN THE STRATA ARE RANDOMLY SELECTED.

Cluster sampling: the population is divided into groups called clusters by some means such as geographic area or schools, etc. Then the researcher randomly selects some of these clusters AND USES ALL MEMBERS OF THE SELECTED CLUSTERS as the subjects of the samples.

Answer 75

It seems like this is almost a blend of the two.... lol.

Answer 76

I get you; you could argue for both! ;)

Answer 77

I am back

Answer 78

Welcome back!

Answer 79

what I should say the prof. " I will just say this is not a good question (: "

Answer 80

Hang on; let me see if I can get a similar example that is stratified with geography

Answer 81

All the examples I've seen use cluster for geography but I do agree that the definition of ''cluster'' should involve ''all members'' in the cluster like you said @DebbieG

Answer 82

?

Answer 83

It is definitely an imperfect question. I've written my share of those, lol. I think you could argue either stratified or cluster, but it doesn't fit either of those perfectly. :)

Answer 84

20 randomly selected stockholders in each of the 50 United States

we are not choosing state right, since we have total 50 states, 20 stockholders each states, we have 20*50=1000 stockholders

Answer 85

The 50 states are naturally occurring pre-existing groups, that would be one argument for cluster sampling

Answer 86

if it was stratified, the researcher would define the groups

Answer 87

(random, systematic, convenience, stratified, cluster) since it does not say etc. we must choose one of them right?

I found this one on the internet as the answer, but bot sure
Probability Sample (Different state populations, still 20 per state)

Answer 88

I would think the ''all members'' part would mean ''all stockholders'' here not all people in each state

Answer 89

ohhh

Answer 90

Yes, I would think you need to choose one of the given choices
''probability sample'' doesn't actually mean anything; not a standard statistics term

Answer 91

how about this argument

It is a stratified random sample. A random sample is one where the researcher insures (usually through the use of random numbers applied to a list of the entire population) that each member of that population has an equal probability of being selected. It provides greater precision than just a simple random sample and it saves money.

Answer 92

oh, the must mean probability bs nonprobability sampling
All the choices are types of probability sampling except convenience sampling. So you would have to go more specific than that.

Answer 93

what you mean..

Answer 94

Its not a random sample because not all members have an equally likely change of being selected due to the groups. its random selection but not random sampling
Once you split it into groups you can do random selection, but you don't have random sampling, if that makes sense. But you haven't said why it is stratified in your argument.

Answer 95

You don't need the extra stuff about saving money because it isn't relevant to why you choose the sampling method you did based on what they gave you

Answer 96

ready for c