Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
12 Cards in this Set
- Front
- Back
|
Sample Size
|
the identified and selected population subset for the survey
-selected to rep. teh entire population - n= 200 Making inferences about the population -(not to describe the sample) -frequency distributors -measures of cental tendancy -measures of dispersion (range, standard deviation) |
|
|
Sample Size Determination: considerations
|
Budget & cost of sample
collection method variability of the population accuracy: how accurate does the sample estimate need to be # of groups and subgroups w/in the sample |
|
|
Sample Size Determination Methods:
|
1) Budget (all you can afford)
2) Rule of thumb/industry standards 3) anticipating subgroups 4) traditional statistical methods -ex: sample size to estimate a mean |
|
|
Samp. Size Determination: Budget
|
start with availible budget
-(subtract) all non-data collection costs / (divide) remainder by cost per resondent = sample size |
|
|
Samp. Size Determination: Rule of thumb/industry stnds.
|
Past Experience
"Gut" feel 100 per subgroup copy test: 200-300 per ad |
|
|
Samp. Size Determination: Anticipating subgroups
|
how many subgroups do you need to examine?
spliting the sample affects sample size (ex: m and f split again by users and non users, in a chart) |
|
|
Samp. Size Determination: Traditional Statistical Methods
|
ex: sample size to estimate a mean
Zsquared * osquared / Esquared Traditional stat. methods: -Population characteristic: is to be estimated by the sample (ex the mean or u) -Sample Statistic: is used to ESTIMATE the population characteristic |
|
|
Sample Size Determination: Tradition Stat. Methods Ctd.
|
Normal Distribution characteristics:
-continuous distribution that is bell shaped & symmetrical about the mean -the mean, median, & mode are = - about 68% of observations are w/in 1 stnd. dev +/- of the mean. 96% are w/in 2 stnd devs and 99+% are w/in 3 stnd. devs Central Limit Thrm: As samp. size increases, the distribution of sample means of size N, randomly selected, appraoches a normal distribution. |
|
|
Tradtional Stat. Methods: Three pieces of info needed
|
1) an estimate of the population stnd. deviation
2) acceptable level of sampling error 3) desired level of confidence |
|
|
Trad. Stat. Methods: Stnd Dev & Error
|
Stnd. Deviation: a measure of the population dispersion
-square root of the variance -how heterogeneous is the population? Error (precision level): how precise the estimates must be -the diff between the sample mean & population mean -ex: avg price willing to pay is $10 +/- $5 |
|
|
Area under Stnd. Noraml curve for Z values of 1, 2, 3
|
1) 68.26 (%)
2) 95.44 3) 99.74 |
|
|
Traditional State Methods: Samp. size to estimate a mean
|
Sample size is determined by the CONFIDENCE LEVEL desired, the allowable ESTIMATE ERROR & the VARIANCE OF THE POPULATION according to :
n = Zsquared * Osquared / Esquared n-necessary sample size Z= parameter to rep. confidence level. Drawn from stnd. normal table osquared = population variance = (Stnd. dev) squared E=error (precision level) |
