Homework 3

Below are examples of this homework assignment from last year. The first two are examples of what we expect for homework 3 for full credit. Pay close attention to the format and simplicity of some of the answers. Be sure to turn in a copy of the article with the appropriate sections highlighted when you turn in your homework and please, staple you homework before you turn it in. The third and fourth examples illustrate what NOT to do; these assignments did not receive high scores. Comments on these assignments are included in capitalized letters. If you have any questions don't hesitate to e-mail us or stop by during office hours.

Homework 3 template

GOOD EXAMPLE 1

Article:

Title: "Oral Cancer Linked With Wart Virus"

Source: Health News

Date: November 6, 1998

I. Goal

The goal of this study was to evaluate a possible association between oral cancer and antibodies to HPV in a person's blood. (Alternatively, the goal could be stated more broadly as an attempt to identify possible causes of oral cancer by looking for correlates.)

II. The Correlation

1. Quote: "Those with antibodies to HPV in their blood-suggesting past exposure to the virus-were 1.7 times more likely than others to develop oral squamous carcinoma."

2. First Variable: Incidence of carcinoma

3. Second Variable: Exposure to HPV 16 antibodies

4. Control Group: 477 healthy people of same age and gender

5. Treatment Group: 284 patients with oral cancer

III. A Model that Postulates a Causal Relation Between the Two Correlated Variables:

1. HPV causes oral cancer.

IV. A Model that Postulates that the Correlation is Caused by a Third, (or Hidden) Variable:

1. One factor the data control for: Cigarette smoking and alcohol consumption (Mark the text where it is stated that this factor is controlled for.)

2. One factor the data do not control for: Excessive and unprotected sex

3. A model in which this correlation is caused by a third (or hidden) variable:

Not a good answer because it restates the original causal model:  People with HPV tend to have higher number of sex partners and moreunprotected sex than those who don't have HPV, but unprotected sex does not cause oral carcinoma, HPV does.

An acceptable answer: Unprotected sex exposes a person to a virus unrelated to HPV that causes oral cancer.  Unprotected sex also increases incidence of HPV, leading to the correlation between HPV and oral cancer.

GOOD EXAMPLE 2

Article:

Title: "Study Says H.I.V. Tests Misstate Women's Risk"

Source: New York Times

Date: November 6, 1998

I. Goal

Does gender play a role in the progression of H.I.V. and A.I.D.S.?

II. The Correlation

1. Quote: "Women infected with H.I.V. may be at a more advanced stage of infection and at a higher risk of developing A.I.D.S."

2. First Variable: progression of H.I.V. to A.I.D.S.

3. Second Variable: gender - female

4. Control Group: Male drug users infected with H.I.V.

5. Treatment Group: Female drug users infected with H.I.V.

III. A Model that Postulates a Causal Relation Between the Two Correlated Variables:

Being a woman causes H.I.V. to progress faster than than does being a man.

IV. A Model that Postulates that the Correlation is Caused by a Third, (Hidden) Variable:

1. One factor the data control for: Viral Load (Mark the text where it is stated that this factor is controlled for.)

2. One factor the data do not control for: The healthy person before infection

3. The model in which this correlation is caused by a third (or hidden) variable:

Women receive worse health care in society so they are weaker and H.I.V. progresses

faster because of this, not because of gender.

Article:

Title: "Likely Cause Found for 'Phantom Pain'."

Source: NOT GIVEN

Date: NOT GIVEN

I. Goal

To show that the brain rewires after limbs are severed

II. The Correlation

1. Quote: "A key portion of the brain rewires after the nerves to a limb are severed, bring sensations from the face into receptors that once "felt" from the arm or leg."

THIS DOESN'T STATE A CORRELATION, JUST A PROPOSED PROCESS.

2. First Variable: brains of a group of eight monkeys remapped their sensations.

THIS IS NOT A VARIABLE.

1. Second Variable: after the nerves leading to one arm were surgically severed at the spinal cord. THIS ALSO NOT A VARIABLE.
2. Control Group: NOT GIVEN

5. Treatment Group: NOT GIVEN

III. A Model that Postulates a Causal Relation Between the Two Correlated Variables:

Sensations are remapped if limbs are severed.

IV. A Model that Postulates that the Correlation is Caused by a Third, (or Hidden) Variable:

1. One factor the data control for: repetition. THIS IS NOT A VARIABLE EITHER
2. One factor the data do not control for: blind.

THIS ISN'T A SUFFICIENT EXPLANATION OF A LACK OF CONTROLS.

3. The model in which this correlation is caused by a third (or hidden) variable:

Some Silver Spring monkeys feel sensation when limbs are severed.

THIS MODEL IS NOT DIFFERENT THAN THE FIRST ONE GIVEN.

POOR EXAMPLE 2

Article:

Title: "Underweight Babies May Become Obese Adults"

Source: Houston Chronicle

Date: November 3, 1999

I. Goal

The purpose of the study was to find out whether underweight babies have a better chance to

later become obese than healthy babies having normal weight

II. The Correlation

1. Quote: "underweight newborns began life with less muscle than other babies but soon had more fat, relative to their size"

2. First Variable: incidence of obesity

3. Second Variable: weight of newborn

4. Control Group: babies of normal weight for their gestational age

5. Treatment Group: babies of less than normal weight for their gestational age

III. A Model that Postulates a Causal Relation Between the Two Correlated Variables:

Being underweight as an infant causes obesity late in life.

IV. A Model that Postulates that the Correlation is Caused by a Third, (or Hidden) Variable:

1. One factor the data control for: weight at birth.

THIS IS NOT A FACTOR CONTROLLED; THIS IS ONE OF THE VARIABLES INVOLVED IN THE CORRELATION.

1. One factor the data do not control for: NONE GIVEN

3. The model in which this correlation is caused by a third (orhidden) variable:

Overweight infants tend to remain heavy as they grow up, but overweight infancy does

not cause obesity. Being underweight as an infant causes obesity.

THIS MODEL IS NOT DIFFERENT THAN THE ONE GIVEN FOR THE CORRELATION.

Homework 3 template