1. Statistical Experiments What is Experimental Design

2. Does knuckle-cracking cause arthritis in hands? It is well claimed by many that continuous cracking of knuckles could lead to development of arthritis in your finger joints later in life. Could you design an experiment to test this claim?

3. Does knuckle-cracking cause arthritis in hands? Medical doctor Donald L. Unger of Thousand Oaks, California, cracked the knuckles of his left hand at least three times a day for more than sixty years, but he did not crack the knuckles of his right hand. No arthritis or other ailments formed in either hand. His work earned him the 2009 Ig Nobel Prize in Medicine, a parody of the Nobel prize award. He says that he did it to prove it to his mother. Reference: http://nowiknow.com/knuckle-head/

4. Play video

5. Experimental vs Observational Studies Many other observational studies have come to the same conclusion as Dr Unger. However, these observational studies involve taking people and dividing them into two groups: Those who have previously habitually cracked their knuckles, and those who have not, and seeing whether there is an existing difference in their levels of hand impairment. These studies are not experiments, as there has been no intervention. Our experiments must involve a treatment, we cannot simply observe what has already occurred. Eg: it is not an experiment to see if boys can run 100m faster than girls, but it is an experiment to see if they run faster with shoes on versus shoes off. Task: What are some other experiments that we could conduct in order to see what effects running speed?

6. What is Experimental Design? Experimental design includes both Strategies for organizing data collection Data analysis procedures matched to those data collection strategies

7. Why do we Need Experimental Design? Because of variability We wouldn’t need a science of experimental design if If all units (people or other things we are experimenting with) were identical and If all units responded identically to treatments We need experimental design to control variability so that treatment effects can be identified

8. What is an Experiment? In its simplest meaning, a process or study that results in the collection of data, the outcome of which is unknown. For our purposes, an experiment is a study in which a researcher attempts to understand the effect that a variable (an explanatory variable) may have on some phenomenon (the response) by controlling the conditions of the study. In an experiment the researcher controls the conditions by allocating individuals to groups and allocating the value of the explanatory variable to be received by each group. A value of the explanatory variable is called a treatment.

9. What we are concerned with In this topic we are only concerned with experiments where there is intervention. The experimenter needs to change one thing between the two groups being studied.

10. Two Different Types of Experiments 1.Comparison of two independent groups: Compare outcomes between a treatment group and a control group, or Compare two groups with different treatments NOTE: Compares outcomes between a treatment group and a control group. There is only one piece of information for each individual. Each individual is randomly assigned a group.

11. Two Different Types of Experiments 2.Paired Comparisons: (also known as before/after experiment) Compare outcomes between one treatment and another for the same group NOTE: Data is collected for each person or object before and after an intervention or treatment. Outcomes are compared between one treatment and another for the same group. There is two pieces of information for each participant.

12. Causality A statistical experiment investigates causality. The goal is to determine whether a change to an independent variable(this change is called the treatment or intervention) causes a change in the dependent variable. You may need to research your topic and discuss information from previous studies.

13. Experimental Variables If an experiment is conducted we look at two variables: The explanatory variable is being investigated to see what, if any, effect it has on the response variable. Example Experiment: Do Year 12 students doing a Standing jump, jump further if they have a target line or not? Explanatory variable: with a target line vs without a target line Response variable: distance of jump The value of the explanatory variable is called a treatment and is controlled by the person conducting the experiment.

14. Randomisation In a well-designed experiment, the allocation of subjects to groups is done using randomisation. Randomisation attempts to make the characteristics of each group very similar so that if each group was given the same treatment, the groups should respond in a similar way, on average. Ensures that each treatment has the same probability of getting good (or bad) units and thus avoids systematic bias. It justifies one of the key assumptions for the statistical analysis. Our statistical analysis assumes that observations are independent. Randomisation means that our estimates will behave as if they were based on independent observations.

15. Control Group Experiments with independent groups have a control group. This group receives no treatment or receives an existing or established treatment. This allows any differences in the response, on average, between the control group and the other group(s) to be visible. When the groups are similar in all ways apart from the treatment received, then any observed differences in the response (if large enough) among the groups, on average, is said to be caused by the treatment.

16. Experimental Design Principles Consider carefully whether you have maintained controlled conditions for each treatment. Were any individuals able to observe someone else do the experiment before they did it? Ideally, no subject should know anything about the experiment before they do it, as the knowledge might affect how they respond to the treatment.

17. Variability A well-designed experiment attempts to minimise unnecessary variability. By randomly allocating individuals to the two groups you reduce variability, as does larger group sizes. Keeping experimental conditions as constant as possible also restricts variability. For example, if you did the standing jump without a target line with year 12 students, and then did the standing jump with a target line with year 9 students, you have added in another variable (year level), which may / will affect the outcome of the treatment.

18. Replication A single experiment may not necessarily be reliable, therefore to improve the reliability of results, participants can be re-randomised and the experiment repeated i.e. replicated. Consistent results after replication of an experiment improves the confidence that researchers can have in their conclusions. For some experiments, it may be appropriate to carry out repeated measurements. Taking repeated measurements of the response variable for each selected value of the explanatory variable is good experimental practice because it provides insight into the variability of the response variable.

19. Some Types of Experiments 1.Can people memorise words better if they draw/imagine the words as pictures? 2.Does the size of a box affect the weight that people guess it is (even when holding it)? 3.Does using your non-dominant hand affect the speed that you can perform a simple task? 4.Does having your eyes open help you to estimate measurements? 5.Does knowing the dimensions of a page help when drawing dots a particular distance apart?

20. Some Types of Experiments 6.Do people think a drink/food tastes better if they know the brand? 7.Does jumping from your dominant foot improve your jumping length? 8.Does drinking from a bottle/cup labelled 'expired' change the perception of taste? 9.Do you jump further if there is a target line marked out for you? 10.Can you perform a task quicker/better if given some advice? 11.Does practice improve your ____ skill?