**The Scientific Method**

**theory**

**Different Approaches**

case study

Correlation

Descriptive Statistics

Measures of Central Tendency

Inferential Statistics

**Experiment!**

**Research Methods**

**hypothesis**

**operational definition**

**replication**

Sleep boosts memory!

When people are sleep deprived, they have a harder time remembering the previous day's learning.

**sleep deprived?**

made easier (and more accurate) by operational definitions!

survey

naturalistic observation

Phineas Gage

H.M.

Kitty Genovese

Genie

David Reimer

Anna O

in depth research on a SINGLE subject for revealing important general information

taking stock of important facts, environmental factors, and specific behaviors

10 men, 10 women take a survey; survey should reveal the general attitudes of the whole group and specific differences between groups

RANDOM SAMPLE

POPULATION

all possible units--so in the case of a survey, people!

CHS Student Body

a subsection of the total population that accurately represents the diversity of the population

members from all 4 classes, both genders, and different activities

Surveying without random sampling can produce skewed results!

looking at natural behaviors in the real world

can validate other experimental findings

may study topics that are unethical (at least in a lab)

Things like surveys and naturalistic observation often show us that one trait or behavior may be related to another.

the HIGHER the mileage, the LOWER the value

the HIGHER your rate of accidents, the HIGHER your insurance cost

The less sexual content teenagers see on TV, the less likely they are to engage in sex.

The more income rose among a sample of poor families, the lesser the incidence of mental illness among their children.

POSITIVE= rising and falling TOGETHER

NEGATIVE= rising and falling OPPOSITELY

correlational coefficient of -1

correlational coefficient of +1

Just remember that correlation DOES NOT equal causation! Illusory correlations happen all the time.

A football fan believes that every time he wears a specific jersey his team wins, so each time they play, he will only wear that jersey.

A child makes a record amount of goals in a soccer game when wearing his red socks, so he continues to wear his red socks for each future game, believing that the socks are related to his play.

A student fails an exam given on a Monday so he determines that he is unlucky and unable to pass a test if it is administered on future Mondays.

enable a researcher to focus on the possible effects of one or more factors by manipulating some factors and controlling others

random assignment

double-blind procedure

placebo effect

makes experimental groups equivalent and protects the integrity of the experiment

eliminates bias in research results (think drug research; if patients and researchers alike do not know which drug is being administered, the results will have more integrity)

belief that an inert substance is creating positive results

experimental group

control group

independent variable

dependent variable

group exposed to one version of the independent variable

If we're looking at whether breast-fed babies are more intelligent than those not breast-fed, our breast-fed babies would be the experimental group.

provides the basis of comparison

In our previous example, formula-fed babies.

breast milk

the factor being manipulated

the outcome factor

how intelligent the experimental group is v. the control group

**validity**

descriptive research

correlation

experiments

How do you interpret the results?

Measures of Variance

mean

median

mode

range

standard deviation

external validity: does it reflect what would happen in the real world?

internal validity: do we know what caused the results of the study?

i.e. naturalistic observations

experiments

MEASURE!

5 36 36 97 120 247 509

the average measure of how data are distributed; add up and divide the total by how many data points there are

150

5 36 36 97 120 247 509

the middle score--half should be above, half should be below!

shows up most frequently

5 36 36 97 120 247 509

how much numerical ground is being covered

5 36 36 97 120 247 509

504

Work out the Mean (the simple average of the numbers)

Then for each number: subtract the Mean and square the result.

Then work out the mean of those squared differences.

Take the square root of that and we are done!

5 36 36 97 120 247 509

150

21025

12, 996

12, 966

2809

900

9409

128,881

mean= 27,002.3

standard deviation=164.3

a number used to tell how measurements for a group are spread out from the average (mean), or expected value

low standard deviation means that most of the numbers are very close to the average; high standard deviation means that the numbers are spread out.

skewed distribution

examines the lack of symmetry in scores around the average

Normal Curve

even distribution on both sides of the mean

how can you use your data to generalize or know if something is true?

You can conceptualize this visually using a histogram:

Sometimes affected by CONFOUNDING VARIABLES, or variables you cannot control!

You can use standard deviation to compare scores from different distributions using "z scores."

If Clarence scored a 72 on a test with a mean of 80 and a standard deviation of 8, his z score would be -1.

If Maria scored an 84 on the same test, her z score would be +.5

simplifies interpreting their performance results

I don't measure the diameter of individual nails measured at a nail factory--instead, I measure a representative sample and I can make a generalization about the whole population.

Statistics also help us determine if differences (particularly in experiments) occur by chance:

I'm testing the effects of sugar on short term memory.

experimental group:

control group:

can remember 7/15 nouns

can remember 6.9/15 nouns

That's not a big difference!

such a tiny difference that it's difficult to say whether the results occurred thanks to the independent variable or by chance

lots of tests to determine the percentage ( or p score) that indicates whether a result occurred by chance or not--the smaller the p score, the more significant the results. Cut off is p=.05 or a 5% probability results occurred by chance. If the p score is higher, your results are not significant!

But what about experimental ethics?

Take a look at your Stanford Prison Experiment handout.

What potential ethical (related to moral right and wrong) issues do you see?

informed consent

deception

right to withdraw

protection from harm

confidentiality

privacy

do you know what you're getting in to?

are they only withholding what's necessary?

can you get out of it if you want to?

are you protected from physical or psychological suffering?

is it certain that you're not going to be exposed when data is released?

does the experiment respect your privacy?

take the mean, subtract it from the score, divide by the unit of standard deviation