research skills 6 intro to analysis data Flashcards
population
- collection of items under observation/discussion
- can be finite - possible to count its individuals
- infinite - no end to population or uncountable
- can be real or hypothetical
variates and variables
- similar terms but
- variate - a quantity or attribute whose value can change eg. for the variable sex the variates can be male or female
- variable - any characteristic , number or quantity that can be measured or counted. it can take on different values. eg sex is a variable it can be male or female
cases
- an experimental unit from which data in collected
- eg data is collected from 200 students lab score - the case is the 200 students the variable is the score
observations
- is asetof one or more measurements on a single unit ofobservation
- e.g. if we looked at 100 people (49 males and 51 females), and analysed the variable ’sex’, using the possible variates male or female. We would have 49 observations for male. We would have looked at 100 cases.
what are the types of variables ?
quantitative:
* numerical
* continuous
Can take any value within a range, they are continuous on a scale, the values between the figures have meaning and the data can be fragmented into parts
e.g. birth weight of a baby in kg and g
* discrete
Discrete variables are specific points on a scale, they might change by steps or jumps, the values between have no meaning, often are whole numbers
e.g. number of children a person has – it cannot be 2.4 children it must be 1, 2,or 3 etc
qualitative
* non numerical
* Nominal
there is no natural order to the categories these variables are assigned to
e.g. degree course or hair colour
- Ordinal
there is a natural order to the categories
e.g. months of the year follow an order, satisfaction scale from 1-10, - Dichotomous
there are only 2 options
e.g. yes / no vote, leave / remain vote.
constant
- A quantity which can assume only one value is called aconstant
- They can be mathematical constants which do not vary, or they can be categorical constants
- e.g. when analysing students and exams - the time allowed for a test would be a numerical constant.
e.g. when analysing student performance at St George’s the university is the categorical constant.
when do you use the mode ?
when handling categoric data
when do you use the median or mean
when handling quantitative data
mean is influenced by by outliers however median is not
how do you calculate the mean deviation ?
subtract the mean from each data value
- absolute deviation = drop the negative sign
calculate the mean from the absolute deviations
how do you calculate standard deviation
square the deviation
add the sum of the squares
divide the sum by n-1 where n is the number of values
square root