Md2 Block1 Exam Prep Q Research Definitions

Front 1

Categorical variables

Back 1

Gender
Severity of haemophilia (mild/moderate/severe)

Front 2

Gender
Severity of haemophilia (mild/moderate/severe)

Type of variables =

Back 2

Categorical variables

Front 3

Numerical variables

Back 3

-Serum bilirubin level
-Number of episodes of disease in a patient over a year
-Reduction in blood pressure following antihypertensive treatment

Front 4

-Serum bilirubin level
-Number of episodes of disease in a patient over a year
-Reduction in blood pressure following antihypertensive treatment

Back 4

Numerical variables

Front 5

Ordinal variables

Back 5

The data are categorical.
The categories of response are ordered.

Front 6

The data are categorical.
The categories of response are ordered.

Back 6

Ordinal variables

Front 7

Nominal variables

Back 7

Ethnic group.

Front 8

Types of variable

Back 8

A binary categorical variable can be either nominal or ordinal.

Quantitative data occur when the variable takes some numerical value.

Front 9

ASCII data

Back 9

Is the same as a text file.
Can be saved by a spreadsheet package.

Front 10

Is the same as a text file.
Can be saved by a spreadsheet package.

Back 10

ASCII data

Front 11

Coding of variables

Back 11

- Gender.
- Has the patient experienced any side effects with a new drug?
- What was the most severe side effect that the patient has experienced?
- Age of oldest child in a family.

Front 12

- Gender.
- Has the patient experienced any side effects with a new drug?
- What was the most severe side effect that the patient has experienced?
- Age of oldest child in a family.

Back 12

Coding of variables

Front 13

Multi- and single-coded variables

Back 13

A multi-coded variable usually has to be turned into one or more single-coded variables before being analysed.

Front 14

A multi-coded variable usually has to be turned into one or more single-coded variables before being analysed.

Back 14

Multi- and single-coded variables

Front 15

Age at interview

Back 15

Date of birth and date of interview

Front 16

Date of birth and date of interview

Back 16

Age at interview

Front 17

Often the same data can be collected in many different ways. Select the group of variables below that would allow you to most accurately find out the side effects of a particular experimental drug.

Back 17

Date and type of first side effect.
Date and type of second side effect.
Date and type of third side effect etc.

This is the most flexible for future analysis as it allows data on the date and type of event to be specified separately. However, for analysis purposes, it may be necessary to code the events in some way.

Front 18

Date and type of first side effect.
Date and type of second side effect.
Date and type of third side effect etc.

Back 18

Often the same data can be collected in many different ways. Select the group of variables below that would allow you to most accurately find out the side effects of a particular experimental drug.

Front 19

If you spot outliers in your data set:

Back 19

You can use a non-parametric test to analyse the data.
Right answer.
This statement is true. Non-parametric methods are particularly useful when there are outliers in the data set, as non-parametric methods do not make any distributional assumptions.

You can repeat the analysis both including and excluding the outliers, and include the outliers if the results are consistent.

Front 20

You can use a non-parametric test to analyse the data.
Right answer.
This statement is true. Non-parametric methods are particularly useful when there are outliers in the data set, as non-parametric methods do not make any distributional assumptions.

You can repeat the analysis both including and excluding the outliers, and include the outliers if the results are consistent.

Back 20

If you spot outliers in your data set:

Front 21

Checking your data for errors:

Back 21

Can be done by typing the data in twice and making the appropriate comparisons.

Front 22

Can be done by typing the data in twice and making the appropriate comparisons.

Back 22

Checking your data for errors:

Front 23

The following are a list of dates of birth of participants in a study of middle-aged women carried out in 1999. Select all of the dates which you suspect may be erroneous.

Back 23

19/05/78
Right answer.
This woman would only have been 21 at the time of the study so she was not middle-aged.

31/04/62
Right answer.
There are only 30 days in April so there is an error in this date of birth.

06/01/22
Right answer.
This woman would have been 77 in 1999 and so is probably too old to be described as middle-aged.

11/13/48
Wrong answer.
There are only 12 months in a year. It appears that the day and month have been interchanged.

Front 24

A study was carried out to assess the impact of a four-week diet on the weight of a group of women. The following list contains the weights of 10 of these women before and after their diet. Select all of the women whose weights (kg) may be erroneous.

Back 24

Woman 2: Weight before diet 94kg, weight after diet 27kg.
Right answer.
The woman's pre-diet weight is very heavy and it is unlikely that she has lost 57 kg in four weeks. Therefore, it is likely that at least one of these weights is erroneous.

Woman 5: Weight before diet 63kg, weight after diet 68kg.
Right answer.
This woman appears to have gained 5kg in four weeks. It is worth checking these values, although some people do gain weights on diets so these weights may indeed be correct.

Woman 9: Weight before diet 699kg, weight after diet 63kg.
Right answer.
A value of 699 kg before the diet is an error.

Front 25

The following list contains waiting times (days) between first visiting a GP and a hospital appointment. Select all times which you suspect are erroneous.

Back 25

883
Right answer.
This is unlikely to be the case as the time period is over 2 years.

17.3
Right answer.
A decimal is unlikely as we do not measure the time interval in fractions of a day.

999
Right answer.
999 days is almost 3 years and is unlikely to be the waiting time for an appointment. It is more likely that 999 was used as a missing value code, as is often case.

Right answer.
It is very unlikely that a GP could write a referral letter, and the appointment be allocated within one day of the patient visiting the GP. This should be checked.