1) Most of the research findings in psychology are based studies done on graduate students in American Universities. Are the results of these studies fully valid? If not why?

No, they are not fully valid as the sample does not represent the population (as not every American actually studied) so the results are actually sample biased.

2) A class has 10 female students and 20 male students. What is the easiest way to collect a sample of 3 from the class?

1 female and 2 males.

3) A city has 1 million residents. To check whether the city residents are happy with the municipality what is a good sample size to be 95% confident about the result with ±5% accuracy?

385

4) During a flu epidemic 5% of hospital admitted patients died every week. A town has two hospitals, one small and one large. In one of the two hospitals 12% flue patients died in a week. Was it likely to be the large or the small hospital?

The small hospital, as there are less people which shows "small numbers".

5) A hotel chain has four hotels with 40, 60, 100 and 300 beds. Average hotels have how many beds? During a weekend when all the four hotels were full, an average guest stayed with how many other guests?

The skew of this statics influences the Average.

Average hotels have 80 beds (even though 2 of the hotels do not have the capacity).

When all of the four hotels were full, an average guest stayed with 210,4 other guests.

6) In a company, there is 30% acceptance rate for male job applicants but just 15% acceptance rate for female applicants. What do you need to check before concluding that there is a bias against women applicants in that company?

If the company has different sections where the applicants apply to (Simpsons Paradoxon).

7) Right after a car-race movie targeting young audience was released, a city saw 40 car accidents in a week involving young drivers. What can you conclude from it?

"False causality is the mother of all statistical mistakes"

I can not conclude, that the car-race movie caused the car accidents.

There could be a correlation but it does not mean that one causes the other.

8) Number of accidents in a factory increased from 12 in one year to 18 next year. How do you think the factory safety department will use statistics to underplay the accident data to hide their inefficiency?

They could say that they had only 30 accidents in the last two years making it seem average to have 15 per year (if you split it in half) even though the numbers of accidents rose from 12 to 18 which shows a 50% increase in casualties.(Statisculation)

9) Government claims during their administration more foreign investments came in. The opposition claims during this time more people has lost jobs. What are they doing? (Half-relevant-Data)

The Government wants to show that it has done improvements (in money making) while the opposition claims that there is no working improvement as many people lost their jobs.

You could also say that they made more money as more people got fired (and they have to pay less).

So the Government as well as the opposition use half relevant Data.

10) A jewellery shops advertises that they are giving 50% discount on making charges of gold jewellery. What are they doing?

They give you a discount for making jewellery where the price is made up by themselfes. By giving you the discount for making gold jewellery they can change the price of gold so the discount vanishes.

Momo Erhart

--

3