Unit 3 Research methods... Can someone give me a clear explanation

[madi1234]

I'm having trouble with my Research methods...

-Stratified sampling?? I find this real confusing.. And also whats an advantage and disadvantage.

-Each research design, and their advantages and disadvantages.

-Random Allocation?... Is that just randomly allocating participants to control/experimental group?

-What Beneficence, and how is it an ethical principle??

[Glockmeister]

I'm having trouble with my Research methods...

-Stratified sampling?? I find this real confusing.. And also whats an advantage and disadvantage.

Ok, I'm going to explain this using M&Ms.

In a packet of M&Ms you are going to have a variety of colours. For simplicity's sake, I'm going to say there red, yellow and green. Now, let's say that in this packet, there are 40 reds, 10 yellows and 30 greens. If we were just to take a random sample of this packet of M&Ms, you are going to potentially have an over-representation of reds. You don't need to be able to work out the probabilities of this actually happening however.

To prevent this we obviously need to sample the M&Ms having the proportion of the M&Ms in our population (our population being all the M&Ms in our packet). So, in this case, if we wanted a sample of say, 20 M&Ms to test, in our sample we would need to have 10 Red M&Ms, 2.5 Yellow M&Ms and 7.5 Green M&Ms in our sample(You'd have to round up or down if you aren't willing to cut those M&M). That way, your sample is now truly representative of the colours of the M&Ms in the entire packet.

Advantages
- Obviously it allows you to examine sub-cultures within a population. For example, if you do a survey on whether you like computer games, younger people are more likely to like it compared to the elderly,

Disadvantage
-Time and Money really. It is a pain in the arse, as you can imagine, to organise this. For starters, you need to define what your strata is going to be and then you need to find how who big each sub-population is.

Something to note (based on my experience around Psychology)
- Stratified sampling Random Stratified sampling. The Random in Random Stratified sampling refers to the fact that selection of the M&Ms after we have figured out what proportions of the characteristics we have stratified has been chosen randomly. This isn't necessarily the case, although I'm led to be believe it usually is.

-Each research design, and their advantages and disadvantages.

It's been awhile since I've done VCE Psychology, and I don't want to teach you Unit 4 stuff so I might get someone else to do this.

-Random Allocation?... Is that just randomly allocating participants to control/experimental group?

Yeah pretty much, Just remember don't use the words in the definition. So use some something like "where participants were placed in the control and experimental groups by chance" or something like that.

-What Beneficence, and how is it an ethical principle??

One of the annoying things about VCE is that it's usually a few years out of date. The Principle of Beneficence, has been subsumed into a category of ethical pricinple called "Propriety" in the revised edition of the APS Code of Ethics released in 2007 (link here ). So according to Wikipedia

An ethical concept. With Regard to Medical Ethics, this concept involves the physician making the decisions that are best for the patient, without regard to personal gain or the interests of others. "Do good and avoid evil." Beneficence, defined in this way, is one of the four ethical principles along with autonomy, justice, and nonmaleficence explained by Tom Beauchamp and James Childress in their "Principles of Biomedical Ethics".

[madi1234]

You know with stratified sampling.. so its like representative sampling of population, like the representive gender???...
Is that correct ?


And the designs include for unit three: independed groups, matched-particpipant and repeated measure designs..



Also can you explain experimenter effect? is that just when the experimenter can affect results


And also the placebo effect - i cant get a clear defition right.

[AppleThief]

STRATIFIED SAMPLING

You know with stratified sampling.. so its like representative sampling of population, like the representive gender???...
Is that correct ?

Yes, but not only gender.

With stratified sampling, they assort the population of interest into particular strata/groups, and randomly select a sample from each strata in the same proportion as they exist in the population. They may be divided into age, gender, religion, background, etc.

Example
they could be sorted into mutually exclusive groups such as
- year 7-9 males
- year 7-9 females
- year 10-12 males
- year 10-12 females

or

- Christian men
- Christian women
- Buddhist men
- Buddhist women

Now for the proportional allocation bit. If 60% of the population of interest are Christian men, then 60% of the sample selected would be Christian men, for example
- Christian men (60%)
- Christian women (20%)
- Buddhist men (10%)
- Buddhist women (10%)

RESEARCH DESIGNS
Repeated measures
Participants participate in both experimental and control conditions, aiming to eliminate extraneous variables arising from individual participant characteristics

Example
If an experiment was testing the relationship between exercise and levels of depression, a group of participants who do not exercise could be tested on a depression inventory. After this, they could be made to exercise for a week, and redo the inventory (i.e. repeat the activity, this time it is the experimental condition)

Matched participants
Each participant is paired with someone similar to them (e.g. for factors such as intelligence, age, gender, which may affect results). One member is placed in the control group, and one in the experimental. Aims to minimise differences in results due to individual participant characteristics.

Independent groups
Two or more randomly allocated groups studied in different experimental conditions. Basic form: an experimental group which receives the IV, and the control group which does not, and is used to compare the experimental group's results to. This is quick and easy to administer

EXPERIMENTER EFFECT
This is when the experimenter expects a certain result, and therefore collects or interprets data in accordance with their bias. They may also subconsciously indicate how participants should react.

Double blind procedure
This minimises the experimenter effect by leaving the experimenter unaware of which group (i.e. experimental/control) is which. A third party does the allocation

PLACEBO EFFECT
Occurs when someone expects that something (e.g. experimental condition) will cause them to behave in a certain manner and instead of responding to the IV, they respond to their own expectations. For example, ~50% of the effect of some drugs are not caused by the drug

[madi1234]

THANK YOU A BUNCH....

and can you please explain... order effects? i dont get it.

[Glockmeister]

Counterbalancing is a type of experimental design in which all possible orders of presenting the variables are included. For example, if you have two groups of participants (group 1 and group 2) and two levels of an independent variable (level 1 and level 2), you would present one possible order (group 1 gets level 1 while group 2 gets level 2) first and then present the opposite order (group 1 gets level 2 while group 2 gets level 1). This way you can measure the effects in all possible situations. Obviously there are limitations with this procedure as not all studies can be designed this way and as you increase the number of variables, conditions, etc., it just becomes logistically problematic.

Not quite.

The primary reason you would use counterbalancing is to avoid what is known as "order effects". The best way to illustrate this point is to use an example (this is pretty much the case for all the research methods)

Say I got a group of Psychology students together to complete this quiz in how happy they are. The way I intended to test this is to ask participants to complete a battery of tests. Now, if I were to give them all the tests in the same order to everyone, by the time you get to the last test you are either going to get:

- Them answering the question based on knowledge gained from doing the previous tests
- Them going "can't be fucked" and just not answering or doing the quiz properly

This introduces biases into your report, as the results you have gathered does not reflect the "true" thoughts and feelings of the participants.

[madi1234]

THANK YOU !