2.4: Statistics: The Language of Psychological

Animal research that liberates animals is a terrible idea because most animals die after being released.

V isn't easy to resolve.

The deaths of 100 million laboratory animals every year in medical and psychological research aren't worth the cost according to some commentators.
There are legitimate questions about how well animal models of psychological and medical disorders translate to human conditions.
It's not likely that depression is experienced the same way by a rat as it is by a human.
The knowledge gained from animal research on aggression, fear, learning, memory, and related topics is not valid to humans.

The position may be too extreme.

Direct benefits to humans have been contributed to by some animal research.

Many psychological treatments are based on animal research.
Without animal research, we don't know a lot about the brain and how it's related to mental disorders.
There are few or no good alternatives to using animals to answer critical psychological questions.
Without animals, we wouldn't be able to test the safety and effectiveness of many drugs.

We don't know when we should or shouldn't use animals in research.

It's clear that animal research has yielded important insights about the brain and behavior and that psychologists are likely to rely on such research for some time to come.
It's clear that animal researchers have to consider the costs of death and suffering in their research.
The intense controversy surrounding animal research is unlikely to abate anytime soon because reasonable people will inevitably disagree about how to weigh the pros and cons.

We have mostly spared you the gory mathematical details of psychological research.

We haven't talked much about how psychologists analyze their findings.
If you want to impress your friends with a technical term, there's no reason for alarm.
ScienceCartoonsPlus.com is the source of our information.

At the risk of simplification, psychologists use two types of statistics.

They skewed distribution and had a positive skewed distribution.

To maintain our promise of keeping things simple, we'll only discuss two types of statistics.

As we calculate each, follow along with us on the left half of the table.

The mean IQ is simply the total of the five scores divided by five, which is 102, if our sample consists of five people.

The three Ms tell us different things.

The mode average is a measure of central to give us the most meaningful answer.

Imagine that we asked 100 people from the general population to name their top unlucky number, and that 90 of them said 13, 7 of them said 3, and another said 1000.
The mean would be middle score in a data set, which would be misleading.

Mean is affected by one extreme score, so it's the most common score median and mode aren't.

The Range versus the Standard Deviation is what the threeMs give us.

The two number lines show the same data sets but with different measures of central tendency.

The standard deviation in (a) will be smaller because four of the scores are less than 116.

The mode and median are 95, which show the central tendency of the distribu b.

The mean is 87 in both groups of scores.

The second set of scores is spread out more than the first.
We need a way of describing the differences in the two data sets.

The measure of variability consists of highest and lowest scores.

The range in the first set of IQ scores is only 15 while the range in the second set is 125.

There is a data set.

There is a data set.

There is a data set.

Data set A mode is 45, Data set B mode is 58.

A shocking out whether the differences we've observed in extraversion between men and 50% of Americans are below average in IQ, reported a team of women are believable, or if they're just a fluke occurrence in our sample.

At the Annual Meeting of the American Society of psychologists, imagine that we calculated the means for men and women.
The bell researchers from Nonexistent State University administered IQ tests to a curve and the distribution of scores in both men and women approximated them.
The extraversion sample of 6,000 Americans yielded a score of 10 for men and 9.9 for women, but half of them scored below the scale.

Inferential statistics enter the picture there.

To figure out if the difference is for the mean or not.

Depending on the research design, we can use a variety of statistics.

It is assumed that we can determine whether the finding occurred by chance with this minimum level.
It's probably a real difference in our sample that a statistically significant result is believable.

A finding can be significant, but not significant enough to make a difference in the real world.

Sampling size is a major determinant of statistical significance.
The larger the sample size, the more likely a result will be significant.
Virtually all findings will become statistically significant with huge sample sizes.

It is so small that it would be useless for predicting anything.

Many people's eyes glaze over when they see a lot of numbers, so it's easy to fool them with statistical sleight of hand.

Three examples of how people can misuse statistics and how people can use them correctly are provided here.
Our goal is not to encourage you to lie with statistics, but to equip you with scientific thinking skills for spotting statistical abuses.

Ms. Dee Seption is running for reelection.

A new tax plan for everyone in your state is part of her platform.
99 percent of people in your state will receive a tax cut this year, according to the "fine print" in Ms. Seption's plan.

According to Ms. Seption, the tax cut for the richest people is necessary because she gets her biggest campaign contributions from them.

I'm going to vote for Dee Seption.

Ms. Dee Seption has engaged in a not-especially-subtle deception, suggesting that she's aptly named.

The average person in her state will receive a tax cut under her plan.
The mean is misleading because under Seption's plan, virtually everyone in her state will only get a $100 tax cut.

The mean highly unrepresentative of the central tendency is that only the richest of the rich will receive a tax cut.

Dee Seption should have reported the mode and median as measures of central tendency.
Arrest Rates the median and mode are less affected by extreme scores than the mean.

According to a study conducted by Faulty Conclusion, Pancake before meditation, a form of relaxation that originated in East Asian cultures, reduces crime rates.

The towns that teach their citizens to practice tm will experience a transcendental meditation.

He found a small town, Pancake, Iowa, and taught all the citizens to practice.

According to Dr.

Two months after the introduction of Pancake.

At a major conference, he proudly announces that although the arrest rates in Waffle stayed the same, the arrest rates in Pancake plummeted.

The arrest rates in Pancake were very high.

What is wrong with Dr.

Conclusion may have tricked us, or he may have tricked himself.

It's not worth writing home about.

Many researchers still use that type of graph in statistics, but it's a no-no.

Dr. used a truncated line graph.

When the effects were small, the apparent effects appeared huge.
Next time, he should use the full range of possible scores on his graph.

Representation conducts a study to find out if there is a relationship between nationality and drinking patterns.