Data Collection
PART 1: POPULATION & SAMPLES OVERVIEW PART 2: RANDOM SAMPLING
Random sampling is when everyone/ everything in a
Population is the whole set of interest. population has an equal chance at being selected. This
This could be everyone living in a town means there is no bias.
or all of the items in a delivery etc.
Types of random sampling
A census asks everyone in the (1) Simple random sampling.
population - Uses a sampling frame
- Very accurate - Every one/thing is given a number and then numbers are
- Can destroy all items selected at random.
- Expensive - No bias
- Time consuming - Not suitable for large numbers
A sample asks a select few of the (2) Systematic sampling
populations, in an e ort to gain - Subjects are chosen at regular intervals
knowledge about the population as a - (eg. If 5 are needed from 100 then every 20th is selected as
whole. 100/5=20)
- Quicker - Suitable for large populations
- Less data to collect - Possibility of bias as sampling frame isn’t random
- Cheaper
- Not as accurate (3) Strati ed sampling
Size of a sample a ects validity - Population is split into groups and selected randomly from
People/objects used in a sample are these groups (eg. male and female, age groups etc)
known as sampling units - Very accurate
- Guaranteed to represent the population
- Must be clearly divided and classi ed accurately
PART 3: TYPES OF DATA
Quantitive vs Qualitative
- Uses numerical values - Non-numerical values
- eg. Shoe size, test scores - eg. Hair colour,
occupation.
Continuous vs Discrete
- Can take any value in a range - Has to be a speci c number
- Eg. Time, length - eg. Number of people
Data Tables
- Frequency tables are used to display large amounts of data and
grouped data
- Grouped data means we don’t know the exact values of the data,
they could be anywhere in the range
- Midpoint is the average of the boundaries
- Class width is the range of the class boundaries
fi ffff fi fi
PART 1: POPULATION & SAMPLES OVERVIEW PART 2: RANDOM SAMPLING
Random sampling is when everyone/ everything in a
Population is the whole set of interest. population has an equal chance at being selected. This
This could be everyone living in a town means there is no bias.
or all of the items in a delivery etc.
Types of random sampling
A census asks everyone in the (1) Simple random sampling.
population - Uses a sampling frame
- Very accurate - Every one/thing is given a number and then numbers are
- Can destroy all items selected at random.
- Expensive - No bias
- Time consuming - Not suitable for large numbers
A sample asks a select few of the (2) Systematic sampling
populations, in an e ort to gain - Subjects are chosen at regular intervals
knowledge about the population as a - (eg. If 5 are needed from 100 then every 20th is selected as
whole. 100/5=20)
- Quicker - Suitable for large populations
- Less data to collect - Possibility of bias as sampling frame isn’t random
- Cheaper
- Not as accurate (3) Strati ed sampling
Size of a sample a ects validity - Population is split into groups and selected randomly from
People/objects used in a sample are these groups (eg. male and female, age groups etc)
known as sampling units - Very accurate
- Guaranteed to represent the population
- Must be clearly divided and classi ed accurately
PART 3: TYPES OF DATA
Quantitive vs Qualitative
- Uses numerical values - Non-numerical values
- eg. Shoe size, test scores - eg. Hair colour,
occupation.
Continuous vs Discrete
- Can take any value in a range - Has to be a speci c number
- Eg. Time, length - eg. Number of people
Data Tables
- Frequency tables are used to display large amounts of data and
grouped data
- Grouped data means we don’t know the exact values of the data,
they could be anywhere in the range
- Midpoint is the average of the boundaries
- Class width is the range of the class boundaries
fi ffff fi fi
