STATISTICS: Types of data
TYPES OF DATA
Data can be:
quantitative - numerical data e.g. height of people
qualitatitive - worded data e.g. colour of hair
There are two types of quantitative data can be:
discrete - is data that can be counted/measured exactly e.g. number of pets
continuous - can take any value within a range (interval) e.g area, temperature
STATISTICS: Collecting data
Primary data - data you have colected yourself.
Advantages: you know how and where it was collected / It is more reliable and accurate
Disadvantages: time consuming, expensive.
Secondary data - data collected by someone else.
Advantages: cheap & easy to obtain.
Disadvantages: might be out of date, have mistakes, don’t know how it was collected.
STATISTICS: Grouping data
When the spread of data is too big we often group the data in a frequency table using class intervals. Grouping data can result in loss of accuracy in both calculations & presentations. But it also, makes it easier to analyse!
Bivariate data are used when you measure 2 varibles which are related things e.g. height & weight of children.
You can put data onto different scales:
→ Categorical scales: give numbers or names to the classes qualitative data so they cn be more easily processed. The numbers are only used for labelling the classes - they don't have any other meaning (groups)
→ Ranked scales: give numbers to the classes, so they can be ordered into a list (put into order)
A census is a official survey of all the members of the population
1) If you want to carry out a census you need to carefully define your population. This means knowing exactly who is in the population, so you can be sure you have surveyed everyone.
2)It is easier to to carry out a census on a small population because, they are fewer members to survey.
- It helps the government plan for the future
- The data you get will be more accurate
- Shows the government the population status of the country
- Takes longer to collect, process, and release data than from a sample
- Very expensive to conduct
When it's not sensible or possible to use a cenus, you have too use sampling.
-Sampling is a cheap and easy alternative to a census.
SAMPLING - a small appropiate number of people or things to be surveyed
- It is quicker, cheaper and more pratical than doing a census of an an entire population.
- You don't have information about every member of the population so it can be less accurate.
Before you choose a sample from a population, you need to make a list or map of all the members of the population - this is called a sample frame.
STATISTICS: Simple random sampling
Simple random sampling - is where everything has an equal chance of being chosen e.g. like a lottery
1) In a simple random sample, you randomly select your dample from the sample frame.
2) It's easy do this sampling with a small, well defined population.
3) To select the sample you need to use random numbers.
Here's how you do it:
Number everything/person, then choose these numbers by placing them in a box and selecting the number randomally.
Advantages: Every member of the population has an equal chance of being selected, so it's completely unbiased.
Disadvantages: It's not always pratical or convenient e.g if the population spread iss over a karge area, the researcher will have to travel a lot.
STATISTICS: Stratified sampling
Stratified sampling is samples taken from every/different catogories.
1) Sometimes the population might be made is of groups or categories that contain members which are similar to each other in some way e.g. gender or age groups
2) In these cases you can use stratified sampling you give the different groups in the sample an amount of representation thats proportional to how big they are in the population - which means big groups get more representation and small group get less.
3) Then you choose the right number from each group at random to make your sample.
ADVANTAGE: If you have easy to define catogories in the population (e2.g. males and females) this is likely to give you a representative sample.
DISADVANTAGE: It can be expensive because, of the extra detail involved.
Sample size for each layer =
Size of whole sample/Size of population × size of layer
STATISTICS: Systematic sampling
Systematic sampling - is every nth number is taken from the sampling frame
1) You can use systematic sampling to generate your sample when the population is very large.
2) It works by choosing a random starting point then taking a sample at regular intervals afterwards
You need a well defined population for systematic sampling.
→First you need o number all the products.
→Next divide your sample frame by sample size
→Then choose a random start number
ADVANTAGES: This should produce an unbaised sample and can be carried out by a machine
DISADVANTAGES: The nth item might coincide with a pattern which will make it biased
STATISTICS: Cluster sampling
1) In larger population, you can pick out smaller groups called clusters - they're oftern based on criteria like location e.g. towns within a city.
2) Cluster sampling is where random sample of clusters is chosen and then, every member in those clusters is included in your sample.
3)The closer the distribution of members within the cluster is to the whole population, the less bias there will be in a study.
4) For this type of sampling you only need a lit of the clusters and the members in the sampled clusters.
ADVANTAGE: It's fairly convenient- can save a lot of travel time when the - population is spread over a large area.
DISADVANTAGE: It's easy to get a biased result e.g. people living in the same street could have similar incomes or employment
STATISTICS: Quota sampling
Quota research - Is oftern used for Market Research.
1) The population is divided up into groups.
2) Then an interviewer is told to interiew a certain number of people from each group.
3) This method of sampling is oftern used in interviews carried out on high streets and the final choice of the sample members is down to the interviewer - so it is not random.
Quick to use and any member of the sample can be replaced by one with the same characteristics.
It can easily be biased, though - the sample chosen depends on the interviewer.
STATISTICS: Convenience sampling
1) The sample is chosen for ease and convenience so it's not random.
2) It's taken from a section of the population present at one particular place and time.
3) Convenience sampling is easy to do but your unlikely to end up with a sample that's representative of the population.
-It's easy to take the sample at a place and time which suits the interviewer.
-Don't need a list of the whole population
-It can be very biased
-There's no attempt to make the sample representative of the population surveyed.
STATISTICS: Make sure questions are...
-Include all choice of answers