PY1103 Lecture Notes - Lecture 4: How To Lie With Statistics, Social Desirability Bias, Leading Question

96 views3 pages

amaranthleopard138

14 Jun 2018

School

James Cook University

Department

Course

PY1103

Professor

Anne Swinbourne

For unlimited access to Class Notes, a Class+ subscription is required.

HOW TO LIE WITH STATISTICS

Population – a group which you wish to talk about in the entirety of that group

Sample – smaller subgroup within the larger group

Statics - Gather a sample and try and say something about the population at large

Biased Sampling

- easiest way to cheat the system

- know what you want to say and go and find the group to demonstrate it

- claim that the group is representative of the entire population

- types of bias

o area bias

o self – selection bias

o leading question bias

o social desirability bias

Area Bias

- claiming that your finding from rural north Queensland apply to urban individuals in

Hobart

- The area of your sample needs to be representative of the study population

- When reading news stories or scientific articles, make sure you verify that there is no

bias in the study. Will these results apply to everyone or everything that the authors

are claiming they can generalise to?

Self Selection Bias

- The persons decision to participate may be related to other factors which affect the

study, making the sample non-representative sample.

- Persons decision to participate relates directly to their interest of the study which

effects the study – where generalisations are not represented of the population at

large

Leading Questions

- Do’t you think nurses should be paid more?

- Should everyone try and save water?

- When did you stop beating your dog?

- Tone of the question determines the type of answer you are likely to get

- Social desirability comes into play

Basic Statistics – measures of central tendency

- The mean: basically the average of the data most frequently used measure of the

central tendency

o To calculate: mean = sum of all scores / number of scores

o Changing the value of any score will change the value of the mean

o The mean is sensitive and unique

- The mode: knowing what is the most frequent score was / may be of interest. Mode

is the most frequent score.

o Relatively insensitive to changes in the data set

find more resources at oneclass.com

Unlock document

This preview shows page 1 of the document.
Unlock all 3 pages and 3 million more documents.

Already have an account? Log in

Document Summary

Population a group which you wish to talk about in the entirety of that group. Sample smaller subgroup within the larger group. Statics - gather a sample and try and say something about the population at large. Area bias claiming that your finding from rural north queensland apply to urban individuals in. The area of your sample needs to be representative of the study population. When reading news stories or scientific articles, make sure you verify that there is no bias in the study. The persons decision to participate may be related to other factors which affect the study, making the sample non-representative sample. Persons decision to participate relates directly to their interest of the study which effects the study where generalisations are not represented of the population at large. Tone of the question determines the type of answer you are likely to get. The mode: knowing what is the most frequent score was / may be of interest.

Related Questions

Part II 1.The California Occupational Mortality study data set was employed to assess mortality data for the years 1979-1981. A 2% sample of employed persons from the 1980 census of California was used. It contains the occupation of each person who had died. This study, like any occupation-based study, is restricted to the work life span as far as age of subjects is concerned. This study used ages 16-64 years. Certain persons, because of their source of work, were excluded as subjects: homemakers, retired persons, students, disabled, military personnel, etc. Only main-stream-type employment was used in this study to establish occupations at risk for heavy alcohol drinking. Using the Table below, answer the following:

1.What occupations are most vulnerable to acquiring cirrhosis of the liver from heavy drinking?

2.List several of the occupations identified by the research studies that are more vulnerable to alcohol-induced, cirrhosis related deaths and that have the highest mortality rates.

3. What are possible confounding variables for the research in occupations and heavy drinking that lead to fatal diseases?

4. Based on available research data, develop and construct a web of causation along with the appropriate decision trees for occupational group-related alcohol deaths.

Table. Alcohol Related Deaths in California, 1979-1981 Cirrhosis 5.5% Digestive organ cancers 5.7% Injuries 13.7% Suicide and homicide 5.% All other causes 64.9% 2. Zika and microcephaly Since May 2015,

Brazil has experienced a significant outbreak of Zika virus. In recent years, Brazilian officials reported an increase in the number of babies born with microcephaly.

1. Briefly describe what we know and what we do not know about Zika virus.

2. What has been done and what should be done to prevent Zika virus epidemic in US and world-wide?

Multiple Choice

5. Which of the following statements describes the major advantages of a randomized clinical trial? a. It avoids observer bias b. It lends itself to ethical justification c. It yields results replicable in other patients d. It rules out self-selection of participants to the different treatment groups d. It enrolls representative patients

6. A survey conducted in England revealed that of 224 families in which there had been a known case of poliomyelitis, 56 maintained parakeets as a family pet. In another British survey, 30 out of 99 poliomyelitis patients questioned kept parakeets. The inference that there is some relationship between the presence of a parakeet in a household and the occurrence of poliomyelitis among household members is a. Correct b. Incorrect because of failure to distinguish between incidence and prevalence c. Incorrect because a proportionate ratio is used when a rate is required to support the inference d. Incorrect because a failure to recognize a possible cohort phenomenon e. Incorrect because there is no control or comparison group.

7. An investigator determines the correlation coefficient between triglyceride levels and degree of atherosclerosis in sampled blood vessels to be +1.67. On the basis of this you would conclude that: a. Triglyceride level is a good predictor of atherosclerosis b. Triglyceride level is not a good predictor of atherosclerosis c. High triglyceride levels cause atherosclerosis d. Atherosclerosis cause high triglyceride levels e. The investigator has incorrectly determined the correlation coefficient

8. A screening test of known sensitivity and specificity is applied to two populations. The prevalence of the disease being screened for is 10% in population A and is 1% in population B. Which of the following is true? a. The percent of all negative tests that have false-negative results is lower in population A than in population B. b. Specificity is lower in population A than in population B. c. Reliability is higher in population A than in population B. d. The percent of all positive tests that have false-positive results is lower in population A than in population B. e. Specificity is higher in population A than in population B.

9. Serum cholesterol levels are obtained for four healthy men. The probability that all will fall below the 10th percentile of the distribution of cholesterol for healthy males is: a. 0.4 b. 1 â (0.1)4 c. (0.1)4 d. (0.9)4 e. Cannot be determined from these data

10. In a diabetes screening program, the screening level for a positive blood sugar level in test 1 is set at 160 mg/dl, and in test 2 at 130 mg/dl. The sensitivity is: a. Greater in test 1 b. Greater in test 2 c. Equal in test 1 and 2 d. Dependent on the size of the population being evaluated

PY1103 Lecture Notes - Lecture 4: How To Lie With Statistics, Social Desirability Bias, Leading Question

Document Summary

Get access

Related Documents

PSYC104 Chapter Notes - Chapter 8: Sampling Error, Sampling Bias, Response Bias

PSYC 310 Chapter Notes - Chapter 5: Adderall, Google Scholar, Impact Factor

STATS 7 Lecture Notes - Lecture 5: Response Bias, Simple Random Sample, Confidence Interval

Related Questions