Data analysis

Three levels of measurements

1. Interval

2. Ordinal

3. Nominal

Nominal data

Takes the form of responses in named categories- do not have an order or scale

Ordinal data

Takes the form of a scale each response is either bigger or smaller than the other. However, the gaps between the points are not measurably equal

Interval data

Takes the form of a scale in which the gaps between the points on the scale are measurably equal

Descriptive statistics

Describes Seth’s of data to inferential statistical which infer whether the data is significant

Two types of descriptive statistics

1. Measures of central tendency

Measures of dispersion

Examples of central tendency Measures

Mode, median and mean

Examples of measures of dispersion

Range and standard deviation

What data is a mode most appropriately used for?

Nominal

Does not take into account any of the scores in anyway- not distorted by outliers and only measure of central tendency that can be used on nominal data

What data is a median most appropriately used for?

Ordinal

Does not take into account the value of the other scores-is not distorted by outliers

What data type is a mean most appropriately used for?

Interval but does work for ordinal

Takes into account the value of the scores- can be distorted by outliers

What measure of central tendency should you use on interval data?

Mean but if there is an outlier, median becomes more appropriate

What type of data is the range most appropriately used for?

Ordinal

Only tells us difference between the highest and lowest score and not the variation within the rest of the scores-is do sorted by outliers and can be used on all apart from nominal data

What type of data is most appropriate for standard deviation?

Interval

Takes all scores into account-less distorted by outliers- only suitable for data that is normally distributed- only ca be used if no outliers, if there is range must be used

Distributions

The frequency of a spread of scores in a data set (histogram) the frequency= y-axis and thing being measured= x-axis

What is the criteria for a normal distribution?

1. Symmetrical frequency graph (bell shaped curve)

2. Mean, median and mode are roughly equal

What causes a skewed distribution

Outliers

Positive skew distribution

High outliers/more low scores than high scores

Negative skew distribution

Low outliers/more high scores than low scores

How do you interpret the mean median and mode on a skewed distribution graph?

Mode→ highest point on graph

Median→ always in middle

Mean→ position tells us if its a positive or negative skew

Order of central tendency on positively skewed graphs

Mode→median→mean

Order of central tendency on negatively skewed distributions

Mean→median→mode

What is found in tables of raw data

Show individual scores prior to analysis

What is found in frequency tables?

Frequency of scores→ ore useful than raw data tables as patterns become evident

What can be found in summary tables?

Descriptive statistics e.g. mean or range

What things must a graph have?

• Title

• Axis labelled- clear and operationalised

Three types of graphs for quantitative data

1. Bar chart

2. Histogram

3. Scatter graph

What is a bar chart used for? How must it be drawn?

Experimental research

IV-x-axis DV=y-axis

Either used to represent frequency or nominal data, categories=x-axis and frequency=y-axis→ columns do not touch and have equal width and spacing

What type of quantitive data is a histogram used to represent and how should it be drawn?

Frequency data e.g. interval or ordinal

Columns touch because each one forms a point on a related scale

Which type of quantitive data is a scatter graph used for?

Raw data

Shows a correlation between two variables