Tải bản đầy đủ
Instrument 2. B: Mental Health Screening Form

Instrument 2. B: Mental Health Screening Form

Tải bản đầy đủ


Designing and Constructing Instruments for Social Research and Evaluation

not diagnostic purposes. However, because medications are often prescribed
and adjusted based on such factors as gender, weight, and age, an instrument
used diagnostically probably would ask about these characteristics.

Guidelines for Using the Mental Health Screening Form-III
The Mental Health Screening Form-III (MHSF-III) was initially designed as a rough
screening device for clients seeking admission to substance abuse treatment
programs. Each MHSF-III question is answered either “yes” or “no.” All questions
reflect the respondent’s entire life history; therefore all questions begin with the
phrase “Have you ever . .”
The preferred mode of administration is for staff members to read each item to the
respondent and get their “yes” and “no” responses. Then, after completing all
18 questions (question 6 has two parts), the staff member should inquire about any
“yes” response by asking “When did this problem first develop?”; “How long did it
last?”; “Did the problem develop before, during, or after you started using
substances?”; and,
“What was happening in your life at that time?” This information can be written
below each item in the space provided. There is additional space for staff member
comments at the bottom of the form.
The MHSF-III can also be given directly to clients for them to complete, providing
they have sufficient reading skills. If there is any doubt about someone’s reading
ability, have the client read the MHSF-II instructions and question number one to the
staff member monitoring this process. If the client can not read and/or comprehend
the questions, the questions must be read and/or explained to him/her.
Whether the MHSF-III is read to a client or s/he reads the questions and responds
on his/her own, the completed MHSF-III should be carefully reviewed by a staff
member to determine how best to use the information. It is strongly recommended
that a qualified mental health specialist be consulted about any “yes” response to
questions 3 through 17. The mental health specialist will determine whether or
not a follow-up, face-to-face interview is needed for a diagnosis and/or treatment
The MHSF-III features a “Total Score” line to reflect the total number of “yes”
responses. The maximum score on the MHSF-III is 18 (question 6 has two parts).
This feature will permit programs to do research and program evaluation on the
mental health-chemical dependence interface for their clients.

c02.indd 48

7/9/07 11:46:43 PM

Instruments and Social Inquiry


The first four questions on the MHSF-III are not unique to any particular diagnosis;
however, questions 5 through 17 reflect symptoms associated with the following
diagnoses/diagnostic categories: Q5, Schizophrenia; Q6, Depressive Disorders; Q7,
Post-Traumatic Stress Disorder; Q8, Phobias; Q9, Intermittent Explosive Disorder;
Q10, Delusional Disorder; Q11, Sexual and Gender Identity Disorders; Q12, Eating
Disorders (Anorexia, Bulimia); Q13, Manic Episode; Q14, Panic Disorder; Q15,
Obsessive-Compulsive Disorder; Q16, Pathological Gambling; Q17, Learning
Disorder and Mental Retardation.
The relationship between the diagnoses/diagnostic categories and the above cited
questions was investigated by having four mental health specialists independently
“select the one MHSF-III question that best matched a list of diagnoses/diagnostic
categories.” All of the mental health specialists matched the questions and
diagnoses/diagnostic categories in the same manner, that is, as we have noted in
the preceding paragraph.
A “yes” response to any of questions 5 through 17 does not, by itself, insure that
a mental health problem exists at this time. A “yes” response raises only the
possibility of a current problem, which is why a consult with a mental health
specialist is strongly recommended.
Mental Health Screening Form-III
Instructions: In this program, we help people with all their problems, not just their
addictions. This commitment includes helping people with emotional problems.
Our staff is ready to help you to deal with any emotional problems you may have,
but we can do this only if we are aware of the problems. Any information you
provide to us on this form will be kept in strict confidence. It will not be released to
any outside person or agency without your permission. If you do not know how
to answer these questions, ask the staff member giving you this form for guidance.
Please note, each item refers to your entire life history, not just your current
situation, this is why each question begins– “Have you ever . .”
1. Have you ever talked to a psychiatrist, psychologist, therapist, social worker, or
counselor about an emotional problem? YES NO
2. Have you ever felt you needed help with your emotional problems, or have you
had people tell you that you should get help for your emotional problems?
3. Have you ever been advised to take medication for anxiety, depression, hearing
voices, or for any other emotional problem? YES NO
4. Have you ever been seen in a psychiatric emergency room or been hospitalized
for psychiatric reasons? YES NO

c02.indd 49

7/9/07 11:46:43 PM


Designing and Constructing Instruments for Social Research and Evaluation

5. Have you ever heard voices no one else could hear or seen objects or things
which others could not see? YES NO
6. (a) Have you ever been depressed for weeks at a time, lost interest or pleasure
in most activities, had trouble concentrating and making decisions, or thought
about killing yourself? YES NO
(b) Did you ever attempt to kill yourself? YES


7. Have you ever had nightmares or flashbacks as a result of being involved in
some traumatic/terrible event? For example, warfare, gang fights, fire, domestic
violence, rape, incest, car accident, being shot or stabbed? YES NO
8. Have you ever experienced any strong fears? For example, of heights, insects,
animals, dirt, attending social events, being in a crowd, being alone, being in
places where it may be hard to escape or get help? YES NO
9. Have you ever given in to an aggressive urge or impulse, on more than one
occasion, that resulted in serious harm to others or led to the destruction of
property? YES NO
10. Have you ever felt that people had something against you, without them
necessarily saying so, or that someone or some group may be trying to
influence your thoughts or behavior? YES NO
11. Have you ever experienced any emotional problems associated with your sexual
interests, your sexual activities, or your choice of sexual partner? YES NO
12. Was there ever a period in your life when you spent a lot of time thinking and
worrying about gaining weight, becoming fat, or controlling your eating?
For example, by repeatedly dieting or fasting, engaging in much exercise to
compensate for binge eating, taking enemas, or forcing yourself to throw up?
13. Have you ever had a period of time when you were so full of energy and your
ideas came very rapidly, when you talked nearly non-stop, when you moved
quickly from one activity to another, when you needed little sleep, and believed
you could do almost anything? YES NO
14. Have you ever had spells or attacks when you suddenly felt anxious, frightened,
uneasy to the extent that you began sweating, your heart began to beat
rapidly, you were shaking or trembling, your stomach was upset, you felt dizzy
or unsteady, as if you would faint? YES NO

c02.indd 50

7/9/07 11:46:43 PM

Instruments and Social Inquiry


15. Have you ever had a persistent, lasting thought or impulse to do something
over and over that caused you considerable distress and interfered with normal
routines, work, or your social relations? Examples would include repeatedly
counting things, checking and rechecking on things you had done, washing
and rewashing your hands, praying, or maintaining a very rigid schedule of
daily activities from which you could not deviate. YES NO
16. Have you ever lost considerable sums of money through gambling or had
problems at work, in school, with your family and friends as a result of your
gambling? YES NO
17. Have you ever been told by teachers, guidance counselors, or others that you
have a special learning problem? YES NO
Print Client’s Name: _____________________________________
Program to which client will be assigned: ___________________
Name of Admissions Counselor: ___________________________

Date: ___________

Reviewer’s Comments:
Total Score: ____________ (each yes ϭ 1 pt.)
Source: Carroll & McGinley, 2000. This material may be reproduced or copied, in its
entirety, without permission.

Key Concepts and Terms

c02.indd 51


naturalistic inquiry

scientific theory

content analysis


social science

focus group



iterative process




7/9/07 11:46:43 PM



In this chapter we will
• Explore the concept of measurement.
• Describe the different levels of measurement.
• Examine the relation between levels of measurement and the construction of
individual items.
In the first chapter we defined an instrument as “a mechanism for measuring
phenomena.” Key to that definition is its reference to measurement. Measurement
provides a systematic process for categorizing and quantifying attributes and characteristics of
the things that we observe, experience, or report. To measure something we assign values,
such as numbers (or symbols), and ensure that there is a logical and systematic
relationship between the values and what they represent as well as between and
among the values themselves (Sarle, 1995). This definition applies to all things
that we want to measure, whether they are physical attributes such as size, weight,
or time or psychological and social attributes such as opinions, attitudes, and
For example, artists’ brushes come in a variety of sizes and shapes. The
sizes of round brushes are numbered from 0000, the very shortest and thinnest,
to 24, the longest and fullest. Each increase in number indicates an increase in
size of about one-thirty-second of an inch in both length and thickness. This

c03.indd 52

7/9/07 11:47:27 PM



simple description tells us a lot about measurement. In this case, brush sizes form
a continuum from small to large, and a numbering system has been created to
categorize size. Because we are using a standard scale, fractions of an inch, we
know there is an equal interval between each brush size. Finally, again because
we are using a standard scale, we can create an instrument, such as a ruler, to
provide consistent measurements.
In this and the next chapter, we will examine the concept of measurement
and the attributes of measurement important in developing an effective instrument. Many of the items that you will create for your instrument will present
a series of values for the respondent or rater to consider and chose from. The
relationship between these values is referred to as the level of measurement.

Levels of Measurement
Standardized systems of measurement have not always existed. In earlier times
the length of a foot could quite literally be the king’s shoe size. Or imagine going
to a marketplace where different merchants used different sets of weights to measure out goods and determine their prices. It was not until the late 1700s that the
metric system, one of the first systematic and standardized approaches to measurement, was introduced. For a system of measurement to become standardized,
there must be agreement on the units of measurement. The unit of measurement
for a kilogram, for example, is a block of platinum and iridium that is kept in a
vault at the Bureau of International Weights and Measures outside of Paris. The
standard for the unit of measurement we call a meter is the distance light travels
in 1/299,792,458 of a second (Strauss, 1995).
A system of measurement embodies a relationship between values and what
they represent, such as the designation of inches and feet to measure height or
distance, of pounds and ounces to measure weight, and minutes and hours to
measure time. From this example you can see that one property of measurement
is that the values should be mutually exclusive; inches cannot be used to measure
time and pounds cannot be used to measure height. However, values can be
combined for measurement, as time and distance, for example, are combined as
a measure of speed. An interesting aspect of measurement is that some attributes
we want to measure, such as time or attitudes and opinions, cannot be directly
observed. Instead, we automatically think of the measurement device, such as a
clock or questionnaire.1
A system of measurement will also feature a relationship between and among
values themselves. For example, the distance, or interval, between the one-inch
and two-inch marks on a yardstick is the same as the one between the seven- and

c03.indd 53

7/9/07 11:47:27 PM


Designing and Constructing Instruments for Social Research and Evaluation

Level of

Numbers are assigned to objects or categories, without having
numerical meaning. That is, numbers are used as labels.
Example: Categories of marital status: (1) married, (2) single,
(3) divorced, (4) separated.


Objects of a set are rank ordered on an operationally defined
attribute. There is no fixed, measurable interval between one
number and another number on the scale.
Example: An intensity scale of 1 ϭ Never, 2 ϭ Sometimes,
3 ϭ Often, 4 ϭ Always.


Numerically equal distances, or interval scales, represent equal
distances among attributes.
Example: Height in feet and inches.


Ratio scales have an absolute zero point. At zero there is a
complete absence of the attribute.
Example: Income in dollars.

eight-inch marks. This also holds true for the markings on a clock, bathroom
scale, or speedometer, and knowing that these values are equally distant helps us
when we change measurement devices, exchanging the hands on an analogue
clock for the screen on a digital clock, for instance.
As we said earlier, the relationship between and among the values in a
set is referred to as level of measurement. Measures are classified into four levels: nominal, ordinal, interval, and ratio (see Table 3.1). Each level contains certain
properties that influence the types of calculations we make with its measures.
Knowledge of the properties associated with each level of measurement
is important for understanding the type of information being gathered by
an instrument and what we can and cannot do when analyzing and interpreting
that information.
Nominal Level
Values at the nominal level of measurement can be named and placed into categories, but they cannot be ordered. For example, we can separate individuals
into groups or categories based on eye color: brown, blue, green, or hazel.

c03.indd 54

7/9/07 11:47:28 PM



We can even assign a number for each eye color: brown ϭ 1, blue ϭ 2, green ϭ
3, and hazel ϭ 4. However, the numbers here serve only as place markers; green
does not have a greater value than blue. Another example of a nominal response
format is the choice between two (dichotomous) values, such as male or female, yes
or no, or true or false.2
A characteristic of a good categorical measure is that the values are
mutually exclusive—they do not overlap. This is obvious for eye color, but
may not be for other categories. For example, in a list of religious preferences
that includes Jewish, Muslim, Protestant, Methodist, and Christian, the choices
are not exclusive because a Methodist is a type of Protestant, and both are
Christians. An individual given these choices would have difficulty selecting
just one.
Because the numbers function only as labels, care must be taken when
analyzing nominal data. As labels, the numbers serve to simplify the process
of categorizing and tallying the results. For example, we can report the
frequency with which these eye colors occur: perhaps 1 (brown) ϭ 20; 2
(blue) ϭ 14; 3 (green) ϭ 10; and 4 (hazel) ϭ 17. The numbers we have associated with the eye colors are used solely for coding; they have no numerical value;
we could just as easily label the colors with letters (A, B, C, and D). It is
important to recognize that here we can order attributes based on the number
of responses, putting brown first because there are twenty occurrences, but that
is different from ordering the attribute—eye color—itself. Therefore we can
report the mode or most frequently appearing eye color, which would be brown,
but we cannot report a mean or average, as we do not have common units
of measure.
Ordinal Level
At the ordinal level of measurement, values can be placed into categories that
are rank ordered or rated along a continuum, such as from high to low.3 For
example, educators might be asked to list the teaching approaches they use
in order of least used to most used. As with nominal level values, the response
alternatives can be labeled with numbers: 1 ϭ least used, 2 ϭ sometimes, 3 ϭ
occasionally, 4 ϭ frequently, and 5 ϭ most used. Although these items have
values that range from low (1) to high (5), it is important to realize that the numbers
do not have equal intervals between them as do the numbers on a yardstick;
that is, an item that is rated a 5 is not five times as “distant” as an item rated
1; the distance, or interval, between 2 and 3 may not be equal to the distance
between 4 and 5. In part, this is because we cannot be sure that all the people

c03.indd 55

7/9/07 11:47:28 PM


Designing and Constructing Instruments for Social Research and Evaluation

using an instrument share the same definition of such terms as never, sometimes,
occasionally, usually, or frequently. For this reason most social scientists classify a
rating response scale as an ordinal measure (for example, Streiner & Norman,
1995; Kane, 1997).
Ordinal data can be ranked, and we can use descriptive statistics, such as
frequency distributions, percentages, and the mode and median; the mean or
average is not an appropriate measure of a response scale for an individual item.
Although we cannot compute the average for individual items, for most instruments
we are interested in collecting aggregate data, and the mean of a set of responses
can be calculated and is often reported. For example, suppose we ask the following question on a customer satisfaction survey and obtain responses from 100

of this
listened to
my concerns.

Not sure

Not at all

Very little


Very much

The first respondent rates the item 3, the second rates it 4, the third rates it
4, and so on. We can then calculate the average of the 100 responses for this item
and for all of the other items in the questionnaire. Let’s suppose that the average
for this item is 3.65. What exactly does this mean? We would probably say that
“on average,” customers were somewhat to very much satisfied with this attribute
of service.
Interval Level
Interval measures have a fixed range, or distance, between one point and another.
Perhaps the best example of the use of the interval level of measurement is a
Celsius or Fahrenheit thermometer, where the interval between 30 and 40 degrees
is equal to the interval between 60 and 70 degrees. Because the intervals between
degrees are equal, a number of analytical operations can be performed, such as
computing the average or mean—in this case a mean temperature. However,
interval level data is not a proportion—in this case, 60 degrees Fahrenheit is not
twice as warm as 30 degrees. Consequently, we should not perform statistical
operations based on proportions or ratios.

c03.indd 56

7/9/07 11:47:28 PM



Ratio Level
Finally, the ratio level of measurement has equal intervals and provides for an
absolute zero point, that is, the complete absence of a value. Examples of ratio
level data are age and income. One person can be twice as old or have three times
the income as another. Certain types of questionnaire items will produce data at the
ratio level, such as an item that attempts to measure the frequency of an occurrence,
asking, for example, “How many times in the past two years have you been
charged with speeding?”




❏ 3



The presence of a zero does not guarantee that a response set is a ratio level
measure. The customer satisfaction question, for example, has a response alternative labeled zero, but the response scale is at the ordinal level of measurement. Additionally, many behavioral and psychometric instruments make use of response
scales like this: 0 ϭ not present, 1 ϭ mild, 2 ϭ moderate, 3 ϭ severe, 4 ϭ very
severe. Although the lowest value is given the rating of zero, these indicators can
only be ranked, and we cannot be sure that there are equal intervals between
them. We could just as easily assign the numbers 1 to 5 to these ratings. In this
case the response scale is at the ordinal level of measurement. Or consider an
item that asks the respondent to indicate the number of times he or she has eaten
at restaurants in the past year: None, 1–5 times, 6–10 times, 11–15 times, 16–20
times, 21 to 25 times. Although the numbers are at equal intervals and there is a
zero point, we have clustered the responses into categories, which can be ranked
at best. Therefore this information would be at the ordinal and not ratio level of

Level of Measurement and Item Construction
Let’s see how the level of measurement influences the construction of items.
Suppose your boss asks you to conduct an inventory of computer equipment in
your work environment. She wants you to determine whether the equipment is
at its assigned location. After thinking about the project you come up with an
instrument that uses the following structure. Fortunately the first three bits of
information (item, ID number, and assigned location) are available on a computer

c03.indd 57

7/9/07 11:47:28 PM


Designing and Constructing Instruments for Social Research and Evaluation





2.5 GHZ


Room 87


85 laser


Room 87



Room 88





Found in
room 88


Because your organization has been thinking about upgrading software, you realize this would also be a good time to solicit information to support the decisionmaking process. So with your boss’s approval you develop another instrument to measure aspects of user satisfaction with the software currently installed. Because this
instrument will be evaluative, many of the items use a rating format:




How would you rate your skill level
in using X-Pro Word Processing

Given your job duties, to what extent
does X-Pro Word Processing software
meet your needs?

Here we see how the purpose of the instrument helps to define how items
are constructed. In the first instrument you only need to categorize the responses,
and so they are at the nominal level of measurement. The second instrument
includes items where the responses are rated and therefore are at the ordinal level
of measurement.
It has been observed that “good measures can help turn abstract ideas
into important, relevant findings, while poor measures render invalid seemingly

c03.indd 58

7/9/07 11:47:29 PM