Inferential Stats, Discussions and Abstracts!!
BATs
Identify which inferential test to use for your experiment
Use the inferential test to decide if your hypothesis can be accepted or rejected
Write a discussion based on your statistics and relate to background research
Write an Abstract for your piece of research
Inferential Statistics
• Allow psychologists to make inferences about the findings of a piece of research when only testing a sample of the world’s population.
• Permit psychologists to work out, for a given probability, whether a pattern in their data has arisen by CHANCE or because there was a real difference/correlation – i.e. the pattern is SIGNIFICANT.
Inferential Statistics• In order to use an Inferential test you need
• A null hypothesis
• An alternative hypothesis
Inferential Statistics• What do we mean by chance?
• Psychologists decide on a probability that they will ‘risk’. They state how certain they are that the findings were not due to chance.
Psychologists use a PROBABILITY of
p < (or equal to) 0.05
There is a 5% possibility that the results occurred by chance
The level of significance
P can vary e.g.1% for drug trials
Type 1 and Type 2 Errors
A 5% significance level is used because..• The degree of uncertainty is not a life or death
matter (medical research uses 1%)• If the level of significance is too high (lenient)
e.g. 10% you may reject a null hypothesis that is actually true. TYPE 1 ERROR
• If level of significance too low (stringent) e.g 1% you may accept the null hypothesis when it is not true. TYPE 2 ERROR
Levels of Measurement
• Before deciding which test to use you need to identify the LEVEL of MEASUREMENT used.
• NOMINAL – data in separate categories e.g. red, blue and green
• ORDINAL – data ordered e,g putting people in order of height. The ‘difference’ between each item is not the same
• INTERVAL – Data using units of equal intervals e.g. counting number of correct answers
• RATIO – There is a true zero point e.g. time taken, distance travelled e.t.c
Observed and Critical Values• TEST STATISTIC – single number
produced after doing particular set of calculations depending on test used e.g rho (Spearman’ correlation test) and U (Mann- Whitney)
• OBSERVED VALUE – Value of the test statistic
• CRITICAL VALUE – value the test stat must reach in order for the null hypothesis to be rejected (these are looked up in a special table for that test)
Some tests are significant if the observed value is equal
to or exceeds the critical value e.g. Spearman’s and
Chi-square.For Mann-Whitney and
Wilcoxon to be significant the observed value should
be less than the critical value
Observed and Critical ValuesTo find the appropriate Critical Value in a
table you need to know …• Degrees of freedom (df) – relates to the
number of ppts in the study (N). Independent groups have 2 values of N – N1 and N2. In Chi Square the df relates to the number of cells used.
• One-tailed or two-tailed test – directional hypothesis = one tailed, non-directional = two tailed
• Significance level – p < (or equal to) 0.05
Which Inferential Statistic?
• Use the text book p286-296 to help you decide which Inferential test is most appropriate for your research
• Note down what test you will use and why!
• Now try the test out on your data
• Was your data significant?
• Can you accept or reject your Hypothesis?
Writing a DiscussionWrite a discussion of your findings. Include the
following …• What did you find out? – specifically mention your
descriptive stats – do they match your hypothesis, were they significant – mention the Inferential test stats
• Do your findings match the original research?• Limitations?• Suggested Modifications?• Implications of research on society/psychology?• What could you do next to develop this research?
Look at an example -p 294-299 of old
Complete Companion A2
AbstractThis is basically a summary at the start of a report to ‘whet the reader’s
appetite’!!• No more than 250 words long, it should concisely
summarise the…1.Aim- The aim was to investigate…… 2.Background-Your hypothesis (null and alternative).3.Design- Independent/Repeated Measures.
Variables.4. Sample- Selection? Girls? Boys.5.Results and statistical conclusions of your study-
Inferential & Descriptive..Main findings.• It should be written in the 3rd person and past
tense, in other words you should not write ‘I found….’instead ‘The experiment was carried out …
Mock next week – Thursday!!
• Revise all you have done on Addictive Behaviour and Research Methods