up!"

 

RECREATION AREA “DAY-USE"
INVESTIGATION TECHNIQUES:
A STUDY OF SURVEY METHODOLOGY
WITHIN MICHIGAN STATE PARKS

Thesis for the Degree of M. S.
MICHIGAN. STATE UNIVERSITY
DOUGLAS MELVIN CRAPO
1969

W4“ .1.‘ 7‘.-

U IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII L _  ;-~ I

THESIS 3 1293 10456

. Y .
. .-- 0‘ t, _ ,_ . |f..‘ t! 6'
6"?" ’2 IV} ‘t‘ ’ 'I ‘1’" lel .1 (504.. 1:.

‘4 ”a'ximl‘wim‘.um. III ' 7 "‘

 

 

 

 

 

 

ABSTRACT

RECREATION AREA "DAY-USE" INVESTIGATION TECHNIQUES:
A STUDY OF SURVEY METHODOLOGY WITHIN
MICHIGAN STATE PARKS
By

Douglas Melvin Crapo

This study was concerned with designing and testing a
data-collection system that would gather accurate information about
recreation area ”day-users. " "Day-users" were considered to be
those individuals who used recreation areas for a single day or a
portion thereof. Since it was impractical to separate "day-users"
from ”campers" in the data-collection stage, information was
gathered from each group and then separated during analysis.

The information collected was in three general categories:
socio-economic characteristics of the users, user "origin" and
travel characteristics, and activity characteristics while in the
area. Limitations imposed upon the data collection system were
that it must be inexpensive, it must yield accurate data, and it must
be implemented into the existing administrative framework of

Michigan State Parks.

Douglas Melvin Crapo

The common methods of data collection were reviewed and
the self-enumerative questionnaire was chosen as being "best, "
considering the limitations outlined above. However, as is the case
with most self—enumerative questionnaires, there is a vital problem
of survey bias resulting from non—response. Essentially, this
study deals with measuring the bias effects of non—response within
a sample, and with testing various methods used to attempt to
reduce non-response.

A careful investigation of sampling techniques and question-
naire construction was conducted. On the basis of this investigation,
an initial questionnaire and dispersal-retrieval system were designed.
The questionnaire had an attractive format and appeared to be brief
(although it did contain a substantial number of questions). Ques-
tionnaire distribution was through hand-out by the contact station
ranger at the area entrance. Retrieval of questionnaires was
accomplished by means of "voluntary deposit" into a collection box
placed near the area exit. A representative sample of all Michigan
parks and recreation areas was then chosen for the testing.

In the ”seasonal" phase of the study, a random systematic
sampling procedure was employed to distribute questionnaires to

I

every nth vehicle entering the 'seasonal" sample parks during the

months of July and August. Of the total number of questionnaires

Douglas Melvin Crapo

handed out, approximately 63 per cent were returned. However,
only 39 per cent were totally completed and were included in the
analysis. Each sample park was then visited by interview teams
and information was obtained from nearly all park users (hereafter
called "parameter data"). The information collected by ”voluntary
deposit” throughout the summer months was then compared to

”parameter data. "

It was found that at the 95 per cent level of con-
fidence the accuracy of the ”voluntary deposit" information was very
high. Travel information (characterized by large standard devia-
tions) was the only data that showed an accuracy problem.

The "experimental" study phase was concerned with identi-
fying and testing numerous variables that might affect the amount of
non-response. Testing of each variable was carried out for at least
two days in all of the "experimental" sample parks.

Each variable was tested individually, and the results of
that particular variable were evident through changes in the ques-
tionnaire response rate. It was found that the ”questionnaire"
which had the greatest amount of response possessed the following
characteristics (ranked in descending order of importance).

(1) The questionnaire was handed out at the entrance to

the recreation area and was collected by means of

"voluntary deposit" in a box at the recreation area

exit.

Douglas Melvin Crapo

(2) A pencil was supplied with the questionnaire in order

to facilitate completion.

(3) The questionnaire was printed on a heavy card-stock.

(4) The color of the questionnaire was, in order of

preference, blue, white, brown, tan, yellow and
orange.

(5) The questions concerned with activity analysis were

constructed so as to be "closed" or "structured. "
The use of a simple chart was most effective.

The small difference in amount of return between the
"regular" length questionnaire and the ”post card" length indicated
that, as long as the questionnaire appeared to be of short length,
response rate was unaffected by amount of content.

The printing of a pre-paid address on one side of a "post
card" type questionnaire did not appear to increase the number of
questionnaires returned when a "voluntary deposit" collection system
was used.

The inclusion of a slotted hole in the questionnaire did not
appear to increase the response rate enough to warrant its addi-
tional expense.

The data collection system designed and tested in this study
may be used as a framework for future investigations of recreation

area users.

RECREATION AREA ”DAY-USE" INVESTIGATION TECHNIQUES:
A STUDY OF SURVEY METHODOLOGY WITHIN

MICHIGAN STATE PARKS

By

Douglas Melvin Crapo

A THE SIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Resource Development

1969

ACKNOWLEDGMENTS

I wish to express my sincere appreciation to all those who
have assisted me with the preparation of this thesis.

I particularly wish to thank Dr. Michael Chubb, thesis
advisor and Director of the Recreation Research and Planning Unit,
for his generous assistance in all phases of the study. His enthu-
siastic encouragement, conscientious guidance and liberal allotment
of financial aid were largely responsible for the success of this
project.

Special thanks also go to Louis F. Twardzik (Chairman,
Department of Park and Recreation Resources), Raleigh Barlowe
(Chairman, Department of Resource Development), and Harry
Webb (Department of Sociology). In addition to their suggestions
and assistance with this study, they have also had a profound influ-
ence upon my professional education and life.

Mr. William Colburne (Division of Recreation Resources
and Planning) and Mr. Paul Rearick (State Parks Division), together
with other staff members from the Michigan Department of Natural

Resources, aided in study design and data collection. The close

ii

cooperation received from all Department of Natural Resource
employees was most gratifying.

Thanks are also extended to:

Dennis Gilliland (Department of Statistics and Probability),
for unending patience and guidance during the statistical design and
analysis phases.

Willard G. Warrington and the staff of the evaluation ser—
vices unit, University College, for assistance in data processing.

Lauren P. Brown, David Marquette and the staff of Univer-
sity Mimeograph and Printing Service, for quality printing under the
most trying circumstances.

John Kohmetscher, Charles Poland, and the staff of Appli-
cations Programming, Computer Laboratory, for computer pro-
gramming and analysis assistance.

Ronald Hodgson, Julia Jones and other graduate students
and student assistants, for unselfish assistance in the design and
implementation of the study.

William Lontz and the staff of the Office of Planning Co-
ordination, Governor' 3 Office, State of Michigan, for assistance in
study design and the arrangement of additional funding.

Lastly, I wish to gratefully acknowledge the assistance of
my wife, Jan. Her untiring assistance and loyal support were a

constant aid during this investigation.

iii

TABLE OF CONTENTS

Page
ACKNOWLEDGMENTS..................ii
LIST OF TABLES
LIST OF FIGURES
Chapter

I. THE PROBLEM AND DEFINITIONS OF
TERMSUSED................1

Introduction
Statement of the Problem
Significance of the Problem
Limitations of the Study
Assumptions
Hypothesis

Sub -Hypothe s e 5
Definitions

II. REVIEW OF METHODS OF DATA
COLLECTION................13

Introduction
Survey Bias
Available Methods of Data Collection
Observation
Use
Advantages
Disadvantages
Characteristics of Data Obtained Through
Questioning

iv

Chapter Page

Personal Interview
Uses
Advantages
Disadvantages
Questionnaires
Uses
Advantages
Disadvantages
Design of the Questionnaire
Correction of Non-Response
Comparison of the Personal Interview
and Questionnaire Survey Methods

III. DESIGN OF A DATA COLLECTION
SYSTEM..................46

Determination of the Desi red Information

Determination of the "Best" Survey
Method

Design of the Survey Instrument and
Dispersal-Retrieval System
Sampling

Choice of a Population Sample

Choice of Sample Parks and Recreation
Areas

Design of the Dispersal-Retrieval System

Questionnaire Design

IV. ACCURACY OF COLLECTED
INFORMATION...............64

Preliminary Preparations
Questionnaire Design
Sample Size
Sampling Method
Implementation of the Continuous Data
Collection System
Collection of Parameter Information
Implementation of Parameter
Information Collection
Treatment of Collected Data
Analysis of Data
Results and Conclusions

Chapter

V. DETERMINATION OF "BEST"
QUESTIONNAIRE AND DISPERSAL—
RETRIEVAL SYSTEM

Determination of Variables Which Might
Influence Questionnaire Response
Means of Choosing "Best" Alternatives
Within Each Variable
Sampling Method
Implementation of the Testing Schedule
Analysis of Data
Results and Conclusions

VI. RECOMMENDATIONS
General Recommendations
Recommendations to the Michigan
State Park System
SELECTED BIBLIOGRAPHY
APPENDICES

Appendix A: "Seasonal" Questionnaire

Appendix B: Weekly Report Form Used
in "Seasonal" Sample Parks .

Appendix C: "Short-Form" Questionnaire
Used as the Interview Schedule

Appendix D: Optical Scan Forms Used in
Coding

Appendix E: Typical "Experimental" Sample
Park Test Instructions

Appendix F: Description of Various
"Experimental" Questionnaires

Appendix G: ”Experimental" Questionnaires
Handed Out, Returned, and Completed

vi

Page

91

115

123

128

130

131

132

134

136

145

LIST OF TAB LE S

Table Page

1. Sample Parks of the Michigan State Park

System Included in This Study . . . . . . . . . 58
2. Number of Questionnaires Handed Out, Returned,

and Judged Complete within ”Seasonal"

SampleParks 83
3. Number of Completed Questionnaires Obtained

from "Seasonal" Sample Park Users

During Parameter Information Collection . . . . 84
4. Variables from "Day-User" Questionnaire

Information Which Show a Significant

Difference in a Comparative Analysis . . . . . 86
5. Variables Which Might Affect Questionnaire

Response and Which Were Capable of Being

Tested in This Study . . . . . . . . . . . . . 93
6. Schedule of "Experimental" Tests . . . . . . . . . 101
7. Results of the Testing of "Experimental"

Variables..................107

vii

Figure

10.

11.

LIST OF FIGURES

Interview Structuring and Interview Depth .

Michigan State Parks and Recreation Areas
Indicating Sample Park Locations

Vehicles Waiting to Enter Dodge Brothers
#4 State Park

Contact Station Rangers Handing Out
Questionnaires

Handing Out Questionnaires While Remaining
Within the Contact Station .

View of Motivational Sign and Pneumatic
Traffic Counter Hose From the

Contact Station

Sign Warning Motorists of Their Approach
to the Deposit Box

Deposit Box and Sign

Pneumatic Traffic Counter Modified for
Systematic Sampling Use

Workers at the Study Center Preparing
Experimental Questionnaires for

Shipment to Parks

Activity Analysis Questions from Ques-
tionnaires in Experimental Test #2

viii

Page

23

58

72

72

73

73

74

74

75

75

141

Figure
12. Back View of "Post Card" Questionnaires
Used in Experimental Tests #5 and #8
13. Front View of ”Post Card” Questionnaires
(2—4-1, 2—4-2, 2-4-3) Used in
Experimental Test #5
14. Front View of "Post Card" Questionnaires

(2-5-1, 2-5-2, 2-5-3) Used in
Experimental Tests #5 and #8

ix

Page

143

144

144

CHAPTER I

THE PROBLEM AND DEFINITIONS

OF TERMS USED

Introduction

 

In early 1968, authorities in the Michigan Department of
Conservation expressed a need for detailed information about the
Michigan State Park "day-user."1 In the past, limited information
about campers had been collected while data on ”day-users” was
virtually nonexistent.

The desired information was of the type that would expose
trends in recreation area usage. These trends would then be used
in the planning process, to ensure that state parks provide the types
of activities and facilities that appear to be most desirable. Since
"trends" are changes over time, collection of trend information must

take place over an extended time period. Thus, the means of

 

A "day-user" is an individual who uses a recreation area
for the day or a portion thereof. He does not reside within the
recreation area during usage.

collecting the desired information would be through a continuous data
collection system.

An investigation into the design of such a system revealed
that there were too many ”unknowns" to be able to both design and
implement a data collection system of the desired scale within one
season. This study represents a compromise from the above
situation--the first step in the eventual implementation of a continu-
ous data collection system throughout all state parks and recreation

areas in Michigan.

Statement of the Problem

 

The problem is to design and test a system to collect
recreation area "day-use" information on a continuous basis within
Michigan State Parks. Specifically, the problem is to design a data
collection system using a questionnaire survey instrument and then
test the system for accuracy of collected information.

The questionnaire survey instrument has many inherent
problems when accurate information is required. Can these prob-
lems be eliminated or at least decreased to the point at which col-

lected information is still of value to administrators?

Significance of the Problem

 

The predicted increase in outdoor recreation participation,

the fixed amount of land available for outdoor recreation facilities,

and the expected growth in population size combine to focus attention
upon a large increase in demand for outdoor recreation opportunities.
”One thing is clear; the conventional approach to providing outdoor
recreation is not adequate for present needs, and it will certainly not
be adequate for the future. "1 These facts point out the importance of
good planning and administration of recreation areas. But adminis-
tration depends upon accurate data to provide an information base for
decision making:

It is a tribute to the recreation leaders of the past that they
were able to accomplish so much without the support of syste-
matic research. . . . As a major activity in contemporary
social life, and as a broad field of public and private enterprise,
outdoor recreation has had relatively little systematic study.

It is generally agreed that much of the research completed

[to date has used questionable survey methods and such low levels of
statistical confidence as to render the collected data suspect. Although
these studies are of value in delineating problem areas and in investi-
gating new survey research techniques, the data collected is often

little better than no data at all. Unfortunately, regardless of past

inaccuracies, recreation survey results have been used in the

decision-making process.

 

1Outdoor Recreation Resources Review Commission, Out-
door Recreation for America (Washington, D. C. : Government
Printing Office, 1962), p. 81.

 

 

2Ibid. , p. 183.

As administrators recognize the value of research, the
demand for accurate data is increased. Sound planning, facility and
program development, budget preparation, and economic analysis
all require basic data about the recreation user and the recreation
site. After the conceptual framework is designed for a particular
problem, and the appropriate information collected and presented in
an effective form, alternative choices can be formulated from the
data and the most appropriate decision can be made. Some of the
specific problems facing recreation agencies in Michigan are out-

lined in Michigan' 5 Recreation Future. 1 Among those mentioned

 

are the lack of funds, the increasing pressure on resources, the
insufficient planning for recreation use, the low level of appreciation
of values of recreation, and the inadequate facilities to meet the
demand.

In the past, the majority of outdoor recreation research has
been centered around the camper. The problems presented by camp-
ground use occupied a disproportionate amount of a recreation
manager' s time. As a result there has been a great deal of interest
in collecting and analyzing information about the camper. He, of all

park visitors, is one of the few with a fixed location for part of his

 

1Michigan Department of Conservation, Michigan' 5 Recrea-
tion Future (East Lansing, Michigan: Michigan Department of Con-
servation, 1966), pp. 16—17.

 

 

visit, and a campground provides a definite population from which
to draw samples. Most campers are gregarious and loquacious,
hence, good research subjects. Indeed, the main problem encoun-
tered in obtaining information from campers is in terminating the
interview.

”Day-users" have frequently been ignored--probably because
of the many problems that they present to survey research. In com-
parison with campers, "day-users" are mobile-~creating problems
in filling the sampling frame if information is to be collected at one
location. Either the information must be obtained at the site or a
record of attendance must be made (i. e. , recording license plate
numbers) so that contacts can be made at a later date. It is the
general impression of recreation authorities that "day-users" show

a higher degree of motivation than do campers. The amount of time

 

available to pursue recreational activities is more limited for the
"day-user”--the recreation day must also include time for travelling
to and from the activity area. Therefore, the ”day-user" is gen-
erally anxious to begin his activities, and during participation he
dislikes being interrupted. At the completion of the activities, he
is usually in a hurry to return to his place of residence.

Yet, in most outdoor recreation areas, ”day-users" com-

prise the majority of recreation visits. In 1967, "day-users"

accounted for 67 per cent of the total park visits within the Michigan
State Park System. Campers comprised the remaining 33 per cent.
There are two principal ways in which outdoor recreation

data can be gathered: Special surveys which are designed for one

 

specific problem or need and are generally short-term studies; and

continuous surveys which are designed to collect data over an ex-

 

tended period of time. Continuous data collection surveys are the
only means of exposing usage trends. By relating trends to changes
in society, predictions about future demand and usage can be made.

Because of its ease of administration, the questionnaire has
been used more than any other survey method. 1 The questionnaire,
however, has several problems associated with it which must be
overcome in order to obtain accurate data from its usage. These
problems are outlined in Chapter II, and several solutions are sug-
gested.

As the volume of information collected increases, it becomes
increasingly more important to put collected data into a form which

permits rapid analysis. Also, the involvement of public departments,

 

1Within the context of this study, "questionnaire" is inter-
preted to mean the printed, self-administered questionnaire.
"Interview" is interpreted to mean the verbal interview in which there
is both an interviewer and respondent. The printed format for the
interview is referred to as the "interview schedule” or "schedule. "

such as utility and transportation agencies, in the planning process
forces recreation planners to put their data into a form that can be
used by the other agencies. The increase in popularity of data
processing/computer types of data analysis and storage is with good
reason. Information in this form can be adapted to a number of uses
and can be made available to a number of nonrelated agencies.

The ultimate choice of what data is required and how it will
be collected rests with the administrator. Administrators and
decision—makers must be cognizant of the relative advantages of data

collection methods, and the limitations of the collected data.

Limitations of the Study

 

The following are limitations of this study:

1. This study will be limited to a representative sample
of Michigan State Parks and Recreation Areas.

2. It will investigate only those individuals entering a
Michigan State Park or Recreation Area by means of an authorized
entrance; specifically, an entrance that is regularly supervised by
Michigan State Park staff.

3. It will be limited to the period July 1, 1968, to Septem-
ber 5, 1968.

4. This study will be conducted within the normal limita-

tions upon recreation area user surveys:

 

a) Although the number of people surveyed is
generally large, the amount of money budgeted for
research is small. A desirable data collection method
maximizes the benefits for all financial inputs.

b) Visits to recreation areas are generally made
with enjoyment as the purpose. Any interruption for
information collection must be made as pleasant and as
short as possible.

c) Information collection must take place as soon
after the experience as possible. This is especially
true when objective recall data is sought. Errors in
recall of information are difficult to evaluate, and such
errors should be decreased by collecting the desired
information immediately upon completion of the activity.

d) It is generally desirable that data collection
systems be made to work within the existing administra-
tive and operational framework of the recreation area.
In Michigan State Parks, the staff are usually operating
near maximum load without the additional burden of a
survey. This is especially important when a continuous
data collection system is being considered.

5. This study will consider only those data collection sys-

tems that will function within the limitations outlined.

Assumptions

 

The following assumptions are made within this study:
1. It is assumed that all Michigan State Parks and Recrea-
tion Areas fall within the classes outlined in the classification of

outdoor recreation areas proposed by Clawson in Land and Water for

 

Recreation.

 

2. It is assumed that previous attendance figures for all
State Parks and Recreation Areas are accurate.

3. It is assumed that through intensive research during
the heaviest use period of the recreation system, a system can be
designed that could be operated throughout the total recreation year.

4. It is assumed that with all the information on hand, a
correct judgment can be made as to which method of data collection
is "best. "

5. It is assumed that when objective questions require
answers of recall, both interview and questionnaire techniques will

produce similar data to the same questions.

Hypothe sis

 

Accurate data about the socio—economic characteristics and

usage patterns of Michigan State Park users can be collected on a

 

1Marion Clawson, Land and Water for Recreation (Chicago:
Rand McNally and Company, 1963), pp. 13-16.

 

10

continuous basis by means of a questionnaire survey instrument and

a random systematic sample.

Sub -Hypothe S e s

 

1. Through a process of trial and selective elimination,
a questionnaire and dispersal-retrieval system can be designed
which will obtain a greater number of usable completed responses
than other questionnaires tested under similar conditions.

2. Questionnaires returned within the proposed dispersal—
retrieval system will contain information that does not differ sig-
nificantly from the parameters of the park -using population of the

Michigan State Park System.

Definitions

 

The following are definitions of terms used within this

study:

accuracy: when dealing with a sample, this term refers to the close-

ness to the actual population parameter.

administrative framework: the overall policy, planning, design and

 

operation relating to a particular area.

M: having the highest degree of desirability, suitability or

advantage. This judgment is made after considering several

11

factors, among them cost, information desired, and the
existing administrative framework within which judgment

is to be made.

contact station: a booth designed to control and regulate the vehicular

 

traffic entering and leaving a specific area.

”day-user": any person entering through the recognized entrance

 

into a Michigan State Park who is not included in the follow-

ing two categories:

1. camper--users who camp within the park
overnight.
2. official and service entrants--individuals

performing a recognized service within the park (1. e. ,
park staff, Coast Guard, police, concession employees,

telephone repairman).

fixed alternative ("closed" or "poll") questions: those questions

 

which offer the respondent a choice among two or more

alternatives.

interview: is a face-to—face interpersonal role situation in which
one person (the interviewer) asks a person being interviewed
(the respondent) questions designed to obtain answers per-

tinent to the purposes of the research problem.

12

open end questions: are those that supply a frame of reference for

 

respondent' 8 answers, but put a minimum of restraint on

the answers and their expression.

repetitive entry: a vehicle entering a particular park two or more

 

times in a single day.

state park or recreation area: any area administered and maintained

 

by a state park and recreation agency for the recreation of

the public.

vehicle: any type of self-propelled, wheeled conveyance (i. e. , car,

bi cycle, motor cycle).

Chapter II will examine the methods of data collection com-
monly used in recreation area user investigation. The uses of each
method, together with its advantages and disadvantages, will be

reviewed.

CHAPTER II

REVIEW OF METHODS OF DATA COLLECTION

Introduction

 

In the past, the particular survey research method used
seemed to be dictated by "availability of subjects and expense rather
than suitability of the technique to the criteria the writer had in mind
for the evaluation. "1 Even present research within the field of out-
door recreation apparently is "dictated more by administrative con-
venience . . . than by meaningfulness of data collected. "

Before the particular method of data collection is chosen,
the researcher and administrator must "weigh" each method and
find the one which most closely fits the requirements of the research

situation. Perhaps a combination of methods would yield the most

 

1W. E. Deming, ”On Errors in Surveys, ” American Soci-
ological Review, 9 (1944), pp. 359-369.

 

 

2Marion Clawson and Jack L. Knetsch, "Recreation Re-
search: Some Basic Analytical Concepts and Suggested Framework
for Research Problems, " Proceedings of the National Conference
on Outdoor Recreation Research. Ann Arbor, Michigan, 1959,
p. 15.

 

 

13

14

accurate information. The steps in deciding which research
technique would be most applicable in a given situation are as
follows:

1. Decide what information is desired.

2. What special characteristics does the desired informa—
tion reflect? Can the information be obtained through solicitation
of objective facts or is the respondent forced to state attitudes and
opinions ?

3. Consider the administrative framework within which
the research is to be attempted. What funds are available? What
period of time is available for the data collection and analysis?
What staff will be available and what is the level of competence of
the staff?

4. Weighing all of the advantages and disadvantages of each
technique, and considering the data collection process within the
administrative framework, decide which technique or which combina—

tion of techniques will best fill the research requirements.

Survey Bias

 

The possible sources of bias within sampling surveys are
numerous and there has been much investigation concerning them.

Deming lists thirteen major types of bias together with an even larger

15

number of possible sources of error. 1 An excellent discussion on
the sources of error in surveys is also presented by Cochran, who
concluded that the potential value of a research project is dependent
upon the care given to selecting, evaluating and controlling sources
of variation. 2 In most surveys, one of the most important bias
sources is due to non—response. A later section within this chapter
will discuss this problem in detail.

Motivation of the subject is extremely important to all sur-
vey research. Unless the subject wants to cooperate and is made to
feel that the information he has to offer is important, the validity of
the collected data is suspect. "When the circumstances in which the
report occurs leave us to suppose that the subject' 3 motivation or
the pressures to which he is exposed are such as to prevent a candid
report, we are not likely to give it much credence. "3 Very little is
known about the motivational process itself and this area is in need

of further research. Cannell and Fowler suggest that there is no

 

1Deming, op. cit., pp. 359—369.

2William G. Cochran, Sampling Techniques (New York:
John Wiley and Sons, 1964)--Chapter 13 deals with all possible error
sources and contains a summary on page 389.

 

3Claire Selltiz, et a1. , Research Methods in Social Rela-
tions (New York: Holt, Rinehart and Winston, 1967), p. 237.

 

16

value in research which does not attempt to interest or motivate the

respondent in the survey. 1

Available Methods
of Data Collection

 

 

Several methods are available for data collection, among
them personal interviews, observation, questionnaires, telephone
surveys, and review of records that have already been established.

Telephone surveys have little to recommend them beyond
speed and, in some cases, low cost. If most respondents are located
in the same general area, as is the case with some regional park
authorities, telephoning might be useful in obtaining information.
However, telephone surveys are usually "limited by possible non-
response, uncooperativeness and by reluctance to answer more than
simple, superficial questions. "2

Review of already collected records usually does not yield
the type of information sought in recreation research. Possible uses
of records are to validate responses which have been collected else-

where, and to act as a source for addresses of individuals to be

 

1Charles F. Cannell and Floyd J. Fowler, "Comparison of
a Self-enumerative Procedure and a Personal Interview: A Validity
Study, " Public Opinion Quarterly, 27 (Summer, 1963), 263.

 

Fred N. Kerlinger, Foundations of Behavioral Research
(New York: Holt, Rinehart and Winston, Inc., 1967), p. 397.

 

17

contacted for future research. Usually outdoor recreation agencies
do not have these records available and must go to a secondary
source to obtain the desired information (i. e. , Department of State).
If, for instance, one wishes to obtain a random sample of
vehicles entering a park and then contact them later by mail or tele-
phone, the process becomes complicated and costly. First, a random
sample must be chosen. Second, the vehicles within the sampling
frame must be identified and the license plates must be recorded.
Third, the license plate numbers must then be forwarded to the De-
partment of State. Fourth, the Department of State identifies the
owner of the vehicle--this process usually takes a period of months.
Fifth, the vehicle owner must then be contacted by mail or telephone
to elicit his cooperation in the research. One must also keep in mind
that the vehicle registrations at the Department of State are not always
up to date and that a certain percentage of the sampling frame will
be lost.
A description of the uses, advantages and disadvantages of

the three most popular methods of survey research follows.

Observation

 

Use

 

Observation reflects the network of actions and reactions

among group members, or the objective behavior of individuals. As

18

a scientific tool, observation can be an accurate means of measuring
actions to the extent that it:

(1) serves a formulated research purpose, (2) is planned

systematically, (3) is recorded systematically and related to

more general propositions rather than being presented as a

set of interesting curiosa, and (4) is subjected to checks and

controls on validity and reliability. 1

The type of observation that is most commonly used in

recreation research is non-participant observation. In this method,
the observer watches and records behavior patterns and does not
attempt to participate in or interfere with the group. This type of
survey research can produce information about how people use the
recreation area. Questions such as "what is the amount of use of an
area (visitor-load) at a specific time?", ”what is the size of the
visitor group?", "what are the activities being participated in?", or
”what is the length of stay?" can be answered through use of the ob-
servation method. The National Parks Service of Canada has made

extensive use of observational techniques in order to obtain data for

the above que stions.

Advantages

 

1. Observation can uncover patterns of action that the

participants themselves do not fully comprehend. An observer is

 

1Selltiz, op. cit., p. 200.

2Specific surveys using the observational method are listed
in the Bibliography of this study.

19

unbiased and can offer an objective interpretation of group
behavior.

2. Behavior is recorded as it occurs. The researcher
does not have to depend upon people' 5 retrospective or anticipatory
reports of their own behavior.

3. Observation, if the subject is unaware that he is being
watched, yields information that is not influenced by anything except
the person's own orientations. Direct questioning of an individual

will often influence his behavior after the questioning situation.

Disadvantages

 

1. Highly structured and methodical observations seem
suitable only for fairly small groups, although an observer (or better,
several observers) can effectively record impressions of the "social
atmosphere” or ”general behavior" of somewhat larger assemblages.
The limitation of lack of funds for most recreation research will
generally ensure that the use of observation within recreation areas
will be confined to large groups, thus reflecting general behavior

only.

 

1Matilda White Riley, Sociological Research: A Case
Approach (New York: Harcourt, Brace and World Incorporated, 1963),
p. 186.

 

2Selltiz, op. cit., p. 201.

20

2. This method is applicable to action taking place only in the
present. It obviously cannot be used to refer to periods
prior to the inception of the research; and to extend obser-
vation into the future in order to trace a gradual develop-
ment of some social system property over time requires
a considerable investment of research time.

This technique is obviously unsuitable for a continuous data collection
system.

3. The areas used for outdoor recreation are usually
large, with many diverse activities. Several observers would be
needed to obtain a general overview of behavior on the entire area.

As we have seen, observational techniques are primarily
'directed toward describing and understanding behavior as it occurs.
They are less effective in providing information about a person' s
beliefs, feelings, motivations, anticipations, perceptions or future
plans; and certainly they provide no information about past behavior
or private behavior. To obtain such information, the interview and
the questionnaire have been used.

Characteristics of Data Obtained
Through Questioning

 

 

Data from questioning reflects the subjective network of
orientations and interpersonal relationships; for example, the under-

lying ideas and feelings of the members, their dispositions to act

 

1Riley, op. cit., p. 186.

21

toward others and to define and evaluate these others in various
ways. It must be kept in mind that responses to questions may be
viewed as resulting from the specific item content of the various
questions and the stylistic orientation of the respondent to the ques-
tioning situation. 1 In questions about actions or factual data the
validity of response is only as good as the recall accuracy of the
respondent. Therefore, answers to questions about actions are
indirect reports of the action, described by the respondent as he
perceived and evaluated it.

In addition to the questioning situation, characteristics of
the respondent himself will also influence response. The motivation
or interest of the subject towards the survey is the most important
factor in the majority of studies. 2 Gergen found that age influenced
response, especially when opinions were desired. Differences in
response of the aged were hypothesized to be due to: lack of
familiarity with issues, lack of interest, more constraint in re-
sponses, more complex opinions than younger people, less capability

of fitting them into forced choice categories, and less willingness to

 

1Kenneth J. Gergen and Kurt W. Bach, "Communication in
the Interview and the Disengaged Respondent, " Public Opinion
Quarterly, 30 (Fall, 1966), 385.

 

 

2Cannell and Fowler, op. cit., p. 261.

22

prov1de answers on topics he is not familiar With. Education of
respondent is also an important influence upon answers to survey

questions.

Personal Interview

 

Interviews may be classified according to the "depth" of

"scale" various aspects of the

information desired. An attempt to
interview is contained in Figure 1. As one would suspect, the
easiest information to obtain is that which is the most "shallow. "
Information about occupation, age, and amount of park usage can

be obtained from a highly structured interview--one which requires
a relatively less-skilled interviewer. At the other end of the "depth"

scale is that information concerned with attitudes, feelings and

motivations.

Interviewing is not as simple a skill as one might first sus-
pect. In order to gather reliable information, all interviewers must
use the same techniques and interview situations. This necessitates
an adequate training program, adequate skilled interviewers, and

supervisory control for all interviewers.

 

1Gergen, op. cit., p. 395.

2Cannell and Fowler, op. cit., p. 261.

 

"Depth"
Level Kind of Information SoughL
la. "Objective Facts"

"Surface"
fl
0‘

 

"Deep"

"Personal" l’acts

Opinions on relatively non-
emotional ”topic

Attitude toward relatively

"ego-important" topic

Feelings (or sentiments)

Motivational analysis
Pre-conscious .

. . unconscious. . .

Ve ri fi cation

of

23

Validity

Interviewer

Skill

Needed

no
hi:
0-:
e3
“.320
sung
we,
DOC/i

Examples or
Characteristics

 

Easier —9

 

6-— More Difficult

Less %

 

6— Greater

More Highly
Structured

 

<———-— Unstructured

 

Occupation. home owner-
ship, residence, car
registration. etc.

Law violations, age.
income. sex life, etc.

What do you think of UN;
of Congo; school-bond issue:
much legislation, etc.

Some relevant aspects:
intensity; interrelations
among attitudes: involv«--
ment; self-concept:
reference groups: ambiva-
lence.

Often have not previously
been clearly verbalized;
ambivalence; appeals to
symbols of consensus.

l. e. , that which "lies
behind" opinions, attitudes,
feelings: personal goals
highly complex relationship
to social goals often not
clear: not too much ve rifled
scientific knowledge avail-
able for ”practical" use.

 

Source: Donald W. Olmsted, from material distributed during Sociology 492, Methods of Social
Research, Michigan State University, East Lansing, April, 1968.

FIGURE 1. -- Interview Structuring and Interview Depth

24

Use

 

The interview has three main purposes. First, it can be

used as an exploratory device to help identify variables, to suggest

 

hypotheses, and to guide other phases of the research. Second, it

can be used as the main instrument of the research. In this case,

 

the interview schedule will include questions designed to measure
the variables of the research. The "Visitor Use Study in New Jersey
State Parks" exemplifies this type of interview usage. 1 Third, the

interview can be used to supplement other data collection techniques;

 

to validate collected information, to follow up unexpected results,
and to further investigate the motivations of respondents and their

reasons for responding as they do.

Advantages

 

1. Since the interviewer and the person interviewed are
both present as the questions are asked and answered, there is
opportunity for greater flexibility in eliciting information. Problems

in question interpretation can be noted and corrected immediately.

 

1New Jersey Department of Conservation and Economic
Development, New Jersey State Park Visitor Use Survey 1966-1967
(Trenton, New Jersey: Department of Conservation and Economic
Development, 1968).

 

25

2. The interviewer has the opportunity to observe both the
subject and the interview environment. Subjective evaluations can
be made without the subject being aware.

3. The personal interview usually yields a high percentage
of returns, for it is difficult for people to refuse the interviewer.

4. It can be made to yield an almost perfect sample of the
general population because practically everyone can be reached by
and can respond to this approach.

5. The information secured is likely to be more correct
than that secured by other techniques when "feelings" or motivations
are sought, since the interviewer can clear up seemingly inaccurate
answers by explaining the questions to the informant. If the latter
deliberately falsifies replies, the interviewer may be trained to spot
such cases and use special devices to get the truth. 2

6. Visual material to which the informant is to react can
be presented.

7. Return visits to complete items on the schedule or to
correct mistakes can usually be made without annoying the informant.
Thus greater numbers of usable returns are assured than when other

methods are employed.

 

1Selltiz, op. cit., p. 238.

2Mildred Parten, Surveys, Polls and Samples: Practical
Procedures (New York: Harper and Brothers, 1950), p. 80.

 

 

26

8. The interviewer may catch the informant off guard and
thus secure more spontaneous reactions than would be the case if a
written form were mailed out for the informant to mull over.

9. The interviewer can usually control which person or per-
sons answer the questions, whereas in mail surveys several members
of the household may confer before the questions are answered. Group
discussions can be held with the personal interview method if desired.

10. The personal interview may take long enough to allow
the informant to become oriented to the topic under investigation.
Thus, recall of relevant material is facilitated.

11. In unstructured interviews, questions about which the
informant is likely to be sensitive can be carefully "sandwiched in"
by the interviewer. By observing the informant' s reactions, the
investigator can change the subject if necessary or explain the sur-
vey problem further if it appears that the respondent is about to
rebel. In other words, a delicate situation can usually be handled
more effectively by a personal interview than by other techniques.

12. Also, in unstructured interviews the language of the
survey can be adapted to the ability or educational level of the
person interviewed. Therefore it is comparatively easy to avoid

misinterpretations or misleading questions.

 

1Ibid.

 

27

Disadvantages

 

1. Compared to the observational method, the degree to
which one can evaluate or predict behavior from interview data is,
at best, limited. 1

2. The primary weakness is the interviewer himself.
Unless the interviewer is competent and effective the reliability of
data cannot be guaranteed.

3. The transportation costs and the time required to cover
addresses in a large area may make the personal interview method
less desirable than other data collection methods.

4. The human equation may distort the returns. If an inter-
viewer has a certain economic bias, for example, he may uncon-
sciously ask the questions so as to secure confirmation of his views.
In opinion studies especially, such biases may operate. To prevent
such coloring of questions, most opinion surveyors instruct their
interviewers to ask the question exactly as printed on the schedule.

5. The organization required for selecting, training, and
supervising a field staff is more complex than that needed for sur—

veys conducted by other methods.

 

1Selltiz, op. cit., p. 201.
2Jackson, op. cit., p. 11.

3Parten, op. cit., p. 81.

28

6. It is usually claimed that costs per interview are
higher when field investigators are employed than when telephone
or mail surveys are used. This may not be true if the area to be
covered is not too great. If the general public in a community is to
be surveyed the costs of securing a representative sample by tele—
phone or mail inquiries will probably equal or exceed the cost by the
personal interview method, since in the end personal follow -up will
be necessary to round out the sample.

7. If the interview is conducted in the home during the
day, the majority of the informants will be housewives. If a response
is to be obtained from a male member of the household, most of the
field work will have to be done in the evening or on weekends. Since
only an hour or two can be used for evening interviewing, the per-
sonal interview method requires a large staff for studies requiring

contacts with the working population. 2

Questionnaires

 

More studies have been made with a questionnaire than any
other instrument in the field of social investigation. However, as

with all survey methods, questionnaires are impositions--to justify

 

1Ibid.

 

2
Ibid. , p. 82.

29

this imposition and to convince the individual that he should fill out
the questionnaire, resistance must be anticipated and procedures
must be designed to overcome it.

Although questionnaire survey methods are generally grouped
together for discussion, there are many differences among question-
naire techniques which affect response. It has been found that the
proportion replying to a questionnaire depends upon: (1) the popula-

tion being surveyed--those individuals with a higher economic status

 

and more education tend to have a higher rate of response than

others;1 (2) the subject of the survgnas the interest of the respon-

 

dent increases, so does the response rate. 2 Strong agreement or
disagreement with the propositions about which respondents are being
surveyed also increases response rate. 3 Subjects respond to ques-
tions that call for factual answers more often than to those asking

for evaluations; 4 (3) the sponsorship of the survey--”A survey with

 

official backing will normally get a bigger response than one

eminating from, say, a university or a research agency. "5 If the

 

1C. A. Moser, Survey Methods in Social Investigation

(London: Heinemann Educational Books Limited, 1958), p. 181.

 

2Ibid.

 

3Ibid., p. 179.
4 .
Jackson, op. c1t., p. 110.

5Moser, op. cit., p. 179.

30

sponsoring body is in some way connected with the population, this

is the most favorable survey situation; (4) the questionnaire length;1

 

(5) the attractiveness of the questionnaire;2 (6) the ease with which

 

 

the questionnaire can be completed and returned. It is apparent that

 

all means must be employed to motivate the subject and to facilitate
completion and return of the questionnaire instrument.

Rewards have been investigated as a means of increasing
response. A gift of a carborundum stone was made to respondents
in a continuing study of recreation mobility in Ontario. It is pre-
sumed that the gift aided in eliciting the desired cooperation. 3 In
1940, Hancock compared the effects of rewards on questionnaire
return in a study of attitudes toward a retail store. The writer in
this case established four groups: the first group received only a
mailed questionnaire; the second, the questionnaire and twenty—five
cents; the third group of subjects received the questionnaire and the
promise that they would be sent twenty-five cents as soon as their
completed questionnaire was received by the investigators; the sub-

jects of the fourth group were interviewed with the same questionnaire

 

1Selltiz, op. cit., p. 241.
2Parten, op. cit., p. 391.

3Ontario Department of Lands and Forests, Ontario Park
User Survey-—Methodology and Results (Toronto: Ontario Depart-
ment of Lands and Forests, 1967), p. A—4, Appendix A.

 

 

31

schedule as used in the mail follow -up. For standardization, the
interviewer asked the question just as it appeared on the instrument
and then marked the form according to the response of the inter-
viewee. The interview was much more productive in terms of the
per cent of completion. The questionnaire alone led to a 9. 56 per
cent return; the questionnaire and quarter brought in a 47. 2 per cent
response; the questionnaire and promise of a quarter 17. 6 per cent;
and the interview return was 85. 5 per cent. 1 However, if the reward
is more than nominal, this may offset the chief advantage of the ques-
tionnaire method--its cheapness.

The manner in which the questionnaire is distributed and
collected is a main influence on amount of return. A questionnaire
survey of campers by Taylor in 1965 received a return of from 60
to 80 per cent when the form was handed out after camper registra-
tion. When the questionnaire was handed out as an integral part of
camper registration the response dropped below 40 per cent.

The amount of response seems to improve as the "captivity"

of the respondent increases. If the subject feels that he has a

 

1J. W. Hancock, "An Experimental Study of Four Methods
of Measuring Unit Cost of Obtaining Attitudes Toward the Retail
Store, " Journal of Applied Psychology, 24 (1940), pp 213-230.

 

2George A. James and Gordon D. Taylor, "Recreation Use
Estimation in the United States and Canada, " XIV IUFRO-Kongress,
section 26 (Munchen, 1967), p. 19.

 

32

definite responsibility to reply, or if he feels he will suffer social
disapproval by not replying, the percentage of response seems to
improve. A difficulty with mail questionnaires, unless they are sent
to a highly motivated group, is that they do not inspire a feeling of

”captivity. "

Uses

The questionnaire can be put to the same uses as the inter-
view, with the exception of an unstructured, in-depth examination of
feelings or motivations. As the problems associated with the ques-
tionnaire technique are decreased, the usage of this technique will

become even more popular.

Advantages

 

1. Several members of a population may be questioned
simultaneously, with no increase in research staff.

2. Questionnaires tend to be less expensive to administer
than interviews.

3. Questionnaires require less skill to administer than do
interviews.

4. The impersonal nature of a questionnaire--the stan-

dardized wording, order of questions, and instructions for recording

 

1Selltiz, op. cit., p. 238.

' 33

responses--can insure uniformity from one measurement situation
to another. 1

5. Respondents have confidence in their anonymity and
thus can feel freer to express views they fear might be disapproved
of.

6. Questionnaires usually place less pressure on the sub-
ject for immediate response. Thus, the subject may ponder a diffi-
cult question before answering.

7. When information concerning several members of a
group or consultation with records is required, questionnaires per-
mit consultation and discussion during the completion.

8. The administration of a questionnaire survey can usually
be adapted so that it doesn' t interfere with the operation of a recrea—
tion area.

9. A mail questionnaire survey can obtain information from
people located in scattered geographical areas.

10. The questionnaire may reach groups who are more or
less protected from solicitors and investigators. In high-rent
apartment houses or private homes where servants protect the occu-

pants from solicitors and other doorbell ringers, for example, it is

 

11bid.

 

2
Cannell and Fowler, op. cit., p. 262.

34

often difficult for investigators to gain admittance. Of course, the
fact that the questionnaire is received at the intended address does
not guarantee that it will be filled in and returned.

11. Personal antagonism to investigators, which may lead
to refusal to give the desired information, is avoided.

12. It is claimed that the mail questionnaire brings many

more returns from the man of the house than does the telephone or

 

personal interview method.

Disadvantages

 

1. The most significant disadvantage is the vital question of
non—response. The effects of failure to fill the sampling frame and
the efforts to overcome this deficiency will be discussed in a later
section.

2. The answers to questionnaires have to be considered as
final. There is no opportunity to probe beyond the given answer, to
clarify an ambiguous one, to overcome unwillingness to answer a
particular question, or to appraise the validity of what the respon-
dent said in the light of how he said it. 2

3. The questionnaire is inappropriate where it is impor—

tant that the views of one person only are obtained, uninfluenced by

 

1Parten, op. cit., p. 95.

2Moser, op. cit., p. 177.

35

discussion with others, and where it is essential that one particular
person in each household or group complete the questions. When
socio—economic class is to be evaluated, information is desired
about the ”head of the family" only. It is difficult to ensure that
information collected through questionnaires applies only to the

"head of the family. "1

4. Questionnaires can be considered only when the ques-
tions are sufficiently simple and straight forward to be understood

with the help of printed instructions.

5. Questions cannot be treated as independent, since the

respondent can see all of the questions before answering.

6. There is no opportunity to supplement respondent' 8

answers with observational data, as there is with an interview.

7. Because most people would rather talk than write,

questionnaires must be made very interesting to induce responses.

8. Mail returns from the last third of the respondents
come in slowly; hence the mail survey must be spread over a
relatively long period, if a high percentage of returns is to be se-

cured.

 

1Ibid.

 

2Parten, op. cit., p. 96.

36

Design of the Questionnaire

 

Questionnaire length is particularly important. Numerous
experiments have found that shorter questionnaires are returned
more frequently than lengthy ones.

When planning questionnaires, it is desirable to use a color
which will attract the recipient' s attention. In some marketing
studies, yellow paper was found to have the highest percentage of
returns and pink followed closely, but dark colors were not effective.
The same investigator also found that when one subject was to re-
ceive a series of questionnaires, alternating of lighter colors solici—
ted more returns than the continual use of one color. 1

The arrangement of questions in order of interest also has
an effect upon response. 2 Lacking the personality of an interviewer
to motivate the respondent towards the survey, the questionnaire
must be its own salesman. The first question should be easily
answered and should have human interest appeal. It should be short,
and preferably one that can be answered with a simple yes or no.

An early interest is a good guarantee of completeness of response.

A few simple questions at the beginning will give the respondent

 

1Ibid. , p. 161.

2Jackson, op. cit., p. 9.

 

37

confidence in himself which is invaluable for carrying the subject
through later more complicated questioning. After rapport has
been established, a few questions asking for the respondent' 5 advice
are good for bolstering the ego and maintaining interest. 1 "General
questions should precede specific ones, otherwise the respondent
is likely to answer the general ones in terms of the earlier specific
ones.”

After questions are assigned specific wording, the follow—
ing points must be checked:

1. Are questions specific enough? Do they ask exactly
what is required?

2. Is simple language used? Any words that cannot be
understood by the desired population must be avoided.

3. Ambiguity must be avoided. A response must mean
one thing only.

4. The use of vague questions encourages vague answers.

5. Avoid leading questions. Wording should be entirely
neutral and should not encourage the respondent to take a position

in his answer.

 

1Parten, op. cit., p. 391.

2
Ibid. , p. 214.

38

6. Hypothetical questions will encourage an answer that
is more a matter of philosophy or wish than fact.

7. One must be sure that he includes all answer possi-
bilities in designing a precoded question. Respondents should not
be forced into an answer category in which they do not belong.

Whether or not one should use open end questions or fixed
alternative questions depends upon the type of answer desired.
Colfax found that the open end form resulted in more complete and
unbiased reports from high school students concerning socio-
economic questions. 1 However, response is higher for fixed alter-
native questions than it is for open-ended questions. 2 Selltiz sums
up advantages and disadvantages of both forms as follows:

Fixed alternative questions have the advantages of being
"standardizable, " simple to administer, quick and relatively
inexpensive to analyze. . . . A closed question may help to
insure that the answers are given in a frame of reference that
is relevant to the purpose of the inquiry and in a form that is
usable in the analysis. . . . Sometimes the provision of
alternative replies helps to make clear the meaning of the ques-
tion. . . . A similar function of alternative responses is to make
clear the dimension along which answers are sought. . . . Fi-
nally the closed question may require the respondent himself

to make a judgment about his attitude rather than leaving this
up to the interviewer or quoter.

 

1David J. Colfax and Irving L. Allen, ”Pre—coded vs.
Open-ended Items and Children' s Reports of Father' s Occupation, "
Sociology of Education, IL (January, 1967), 965.

 

2Jackson, op. cit., p. 110.

39

Closed questions are more efficient where the possible
alternative replies are known, limited in number, and clear
cut. Thus they are appropriate for securing factual informa-
tion (age, education, home ownership, amount of rent, et
cetera) and for eliciting expressions of opinion about issues on
which people hold clear opinions.

Open-ended questions are called for when the issue is com-
plex, when the relevant dimensions are not known, or when the
interest of the research lies in the exploration of a process or
of the individual' 5 formulation of an issue. The closed question
has the advantage of focusing the respondent' s attention on the
dimension of the problem in which the investigator is interested;
by the same token, it does not provide information about the
respondent' s own formulation of the issue, the frame of ref-
erence in which he perceives it, the factors that are salient for
him, the motivations that underlie his opinions. When these
matters are the focus of interest open-ended questions are
essential. 1

An excellent description of the steps to be considered in

constructing a questionnaire and interview schedule is contained in

Appendix C of Research Methods in Social Relations by Selltiz.

 

Within this appendix the points to be considered in training inter-

viewers are also outlined.

Correction of Non-Response

 

Non-response is a problem because of the likelihood--

repeatedly confirmed in practice--that people who do not return

questionnaires differ from those who do. The greater the propor-

tion of non-respondents and the extent to which non-respondents

 

1Selltiz, op. cit., p. 262.

40

differ from those who did respond, the greater will be the bias from

non- response .

The most inexpensive means of dealing with non-response
is by reducing it in the first place. The techniques to accomplish
this have been discussed in previous sections. In spite of all efforts,
however, non-response will usually occur in all survey research.
Some individuals perceive survey research as an invasion of privacy
and their hostility is evident in non-response to questionnaires and
interviews. 2 Even though a respondent might want to help in giving
information, he may be unable to do so. Psychologists have pointed
out that we are not aware of many of our most important beliefs and

motivations, and hence cannot report them.

Most survey researchers will further reduce non-response
by use of follow—up waves. Shafer doubled the response to his
camper survey through the use of four follow-up waves. A some-

what comparable mail back questionnaire survey by the Detroit

 

1Moser, op. cit., p. 128.

2Stanley D. Bachrack and Harry M. Scoble, ”Mail Ques—
tionnaire Efficiency: Controlled Reduction of Non-response, ”
Public Opinion Quarterly, XXXI (January, 1967), 267.

 

3Selltiz, op. cit., p. 237.

41

Metropolitan Area Regional Planning Commission obtained the

following completed response:

Initial Response 14.2% of Total Number
Distributed
Additional response following the 19. 3% of total number
lst "follow-up" wave Distributed
Additional response following the 15.7% of Total Number
2nd "follow-up" wave Distributed
Additional response following the 9. 2% of Total Number
3rd "follow -up" wave Distributed
Total 58. 4%

Another survey within the same regional parks obtained 30. 8 per
cent of the sampling frame in the initial response. Two mail
follow -up waves produced an additional 28. 8 per cent of the sam-
pling frame. 2 It seems apparent that the use of follow—up waves
will increase questionnaire response significantly. However,
follow-up waves are difficult to administer, expensive and time

consuming.

 

1Detroit Metropolitan Area Regional Planning Commission,
Home Survey of Regional Recreation Activities (Detroit: Detroit
Metropolitan Area Regional Planning Commission, 1959), p. 19.

 

2Detroit Metropolitan Area Regional Planning Commission,
Park Users Survey: A Survey of Persons Using Eight Selected
Regional Parks in the Detroit Region (Detroit: Detroit Metropolitan
Area Regional Planning Commission, 1959), p. 21.

 

 

42

A low response rate does not necessarily mean the research
is of no value:
. it would be quite wrong to imply that non-response
vitiates the scientific nature of sampling. Mail surveys apart,
it is usually possible to keep non-response down to a reasonable
level and to estimate roughly what biasing effects it may have
upon the results. 1

Statistical adjustments for non-response are outlined in Sampling

Techniques by Cochran.

 

Even with 100 per cent returns, the researcher must be
cautious about accepting the responses from questionnaires or inter-
views. 3 This is especially true when subject matter deals with
social approval issues or issues of status.

Comparison of the Personal Interview and
Questionnaire Survey Methods

 

 

Comparisons of the interview and questionnaire survey
methods can be made on the following basis:

1. Amount of response--numerous studies have shown

 

that the interview method obtains responses from a larger propor-

tion of the sample.

 

1Moser, op. cit., p. 127.
2Cochran, op. cit., p. 71.

3Elwood Lewis Shafer, Jr. , "Effects of Sampling, Location,
Period, and Method on Camper Survey Results" (unpublished Ph. D.
dissertation, State University College of Forestry, Syracuse Uni—
versity, 1966), p. 13.

43

2. Completeness of return--an interviewer can assure

 

that all questions are answered completely.

3. Quality of answer--the presence of the interviewer

 

generally insures that most answers are "usable. ”

4. Cost of method-~unless extensive callbacks are required,

 

the questionnaire method is considerably less expensive.

In a recent study, Shafer compared various methods of ob-
taining Opinions (preferences) from campers. He used four tech-
niques: personal interviews at the site; questionnaires to be
completed and returned at the camping site; questionnaires mailed
to the camper immediately upon completion of the camping experi-
ence; and questionnaires mailed three months following the camping
experience. Shafer used four follow-up waves in order to decrease
the non-response to the questionnaire methods. He concluded that
hand out and mailed survey procedures could be expected to produce
about the same total response regardless of the intensity of sampling
or the location, but sampling periods throughout the summer may have
a Significant influence on return. 1 In comparing other methods with
the interview he found that answers obtained from the hand out method
differed significantly from corresponding answers obtained from the

interview, 19 per cent of the time. The mailed questionnaires

 

1Ibid., p. 168.

44

varied between 12 per cent and 24 per cent from interview results,
depending upon the sample waves included in the analysis. Shafer

concluded that:

. . based on a complete overview of all these results, the
hand out and mailed survey methods-~with three to four waves
for non-respondents—-generally seemed to be reliable for
obtaining information of the type requested in this study
(opinions of the recreation site) if a 10 to 25 percentage point
discrepancy from comparable personal interview values is
acceptable.

However, if only a 10 to 15 percentageggintflffemnfiaL

can be tolerated, then the delayed mailed procedure is recom-
mended, with the possible exception perhaps of questions that
deal with large value categories. 1

Shafer found that the answers for sensitive questions such as income
and expenditure patterns were the same for all methods, implying
that the potential subject matter which can be accurately treated in
future questionnaire surveys seems quite broad and varied. 2 He
states that if interviews are out of the question, the delayed mailed
survey should be used in preference to hand out or immediately
mailed surveys.

In obtaining information about issues involving social

approval or disapproval, Knudson found questionnaire responses to

 

1Ibid., p. 168.

2Ibid., p. 174.

3Ibid., p. 178.

 

45

be more valid than interview responses. 1 The interviews appeared

to provide information that has been revised to an assumed public

standard or standard held by the interviewer. When information is

desired about a person other than the respondent, Cannell and

Fowler found that the most valid answers were obtained by the ques-
2

tionnaire method. These authors also found that:

. in the interview procedure, education of respondent
is the most important variable, and that motivation is a factor
only within educational groups; whereas in self-enumerative
procedures motivation is a more important variable, and edu-
cation of respondent becomes a factor within the "better
motivated" groups of respondents.

When "objective recall" data is desired, with no attitudes or opinions
called for, very little difference is found between responses from
interview and questionnaire methods. 4

The following chapter will discuss preliminary aspects of

study design.

 

1Dean D. Knudson, "Response Differences to Questions on
Sexual Standards: An Interview-Questionnaire Comparison, " Public
Opinion Quarterly, XXXI (January, 1967), 297.

 

2Cannell and Fowler, op. cit., p. 262.
3 .
Ibid., p. 261.

4Ibid.

 

CHAPTER III

DESIGN OF A DATA COLLECTION SYSTEM

Determination of the
Desired Information

 

 

Several meetings were held with representatives from
interested Park and Recreation agencies to determine the type of
information needed to indicate "day-use" trends. It was concluded
that factual data about the "day-user" was required in three general
categories; characteristics of the user, locational and travel char-
acteristics, and activity and use characteristics. The total infor-
mation outlined as being of value to planners and administrators was

as follows:

Characteristics of the User

 

1. Distinguish between the ”day-user" and the camper.
2. Group composition--what kind of group entered the
park ?

3. Ages and sex of all individuals in the group?

46

47

4. What is the socio-economic level of the family of the

driver of the vehicle?

Travel and Locational Characteristics

 

5. What is the "origin" or residence of the user?

6. What distance is the "origin" of the user from the park?

Activity and Use Characteristics

 

7. What ”type" of a visit is this? For example, one day
event or part of a major annual vacation?

8. How often does the subject use the particular park?

9. How often does the subject use other parks? City
parks? County and metropolitan parks? State parks? National
parks?

10. When did the subject enter the particular park?

11. How much time did the user spend in the park?

12. What activities were participated in and for what length
of time?

Determination of the
"Best'' Survey Method

 

 

Authorities of the Michigan State Park system advised that
a continuous data collection system must function within the following

limitations:

48

1. No additional staff would be made available to supple—
ment the present staff within each park and recreation area. During
the summer season, a maximum of two staff members from division
office would be made available for limited periods of time.

2. Since the process of obtaining names and addresses
from vehicle registrations would be too costly and time consuming,
no methods of dispersal—retrieval were to be considered that would
have to use this information source.

3. Data was desired for each park and recreation area.

4. Funds for the research were extremely limited.

5. The data collection system should be put into operation
as soon as possible.

6. Data collected must have a known accuracy that will
permit it to be used in the planning and management processes.

As previously outlined above, all of the desired information
was "objective recall, " that is, no opinions or attitudes are sought.
"Objective recall" data can be obtained through the use of either
personal interview or questionnaire techniques.

Considering the advantages and disadvantages of each
method as outlined in Chapter II, the data required, and the limita-
tions of the Michigan State Park system, the "best" survey technique

was the questionnaire method. The one weakness of this technique

49

was the low rate of response, thus jeopardizing accuracy of infor-
mation. If this deficiency could be surmounted, questionnaires
could provide data of the desired accuracy.

Design of the Survey Instrument and
Dispersal-Retrieval System

 

 

The testing of each sub-hypothesis involved different parks,
sampling methods and administrative procedures. In order to avoid
confusion, the development of the "best" questionnaire and dispersal-
retrieval system (discussed in Chapter V of this report) is hereafter

' Testing collected question-

referred to as the "experimental phase. '
naire data for accuracy (discussed in Chapter IV) is hereafter referred
to as the ”seasonal phase" because the same type of questionnaire was
distributed throughout the entire summer season. Although the test-
ing of each phase was carried out in different parks, both phases
were administered from the same research center during the same
time period.

Before implementing each phase, several common problems

were encountered. The following section applies to both the "experi—

mental" and "seasonal" phases of the study.

Sampling
If one were to survey every user entering every State Park

or recreation area, a fantastic amount of funds (and staff,

50

administrative space, and analysis resources) would be required.
Fortunately, a sample of visitors will yield the desired information
within precise limits of accuracy. The greatest advantage of sam-
pling is that it saves time, labor and money, allowing more "distance"
for the research dollar. If the sampling procedure is correctly
designed and is followed, the research limitations are minimal.

In order to obtain a representative sample, that is, to
ensure that all segments of the population thatuse the area are

' a random method of

sampled according to their relative ”weights, '
sampling should be employed. The criteria of a random sample is
that every subject in the population has an equal chance of being
included in the sample.

A random sample is difficult to obtain within a recreation
area. It cannot be obtained by choosing individuals participating in
certain activities (because responses will be biased towards that
particular activity and other activities will be ignored), or by choos-
ing vehicles within a parking lot (because not all vehicles are
usually vehicles within a parking lot--some vehicles are parked
along roadways, others can be taken to the activity site, i. e. ,

bicycle). Usually the only locations where truly random samples

can be chosen are at entrances or exits.

 

1The disadvantages of sampling are outlined on page 111
in Surveys, Polls and Samples by Parten.

 

51

Random sampling is relatively simple. Either the use of
random number tables or a systematic sampling system will yield
a truly representative sample, assuming that "enough" sampling
periods (of "long enough" duration) are used.

If the area has two or more entrances or exits the task
becomes more complicated. The proportion of total vehicle flow
either into or out of the area must be calculated for each entrance
or exit. A correspondingly proportionate amount of the random
sample must then be obtained for each of the entrances or exits.

For example, if a particular area has three entrance/exits, with
proportions of total vehicular traffic of 20 per cent, 30 per cent and
50 per cent respectively, then 20 per cent of the total sample should
be obtained at the first entrance/exit, et cetera.

When a sample is used to represent the entire population,
the science of statistics can be employed to make inferences about
the population based upon information contained in the sample. Inas-
much as populations are "characterized by numerical descriptive
measures called parameters, statistical inference is concerned with
making inferences about population parameters. "1 Typical population

parameters are the mean, standard deviation, or the area between

 

1William Mendenhall, Introduction to Probability and Sta-
tistics (Belmont: Wadsworth Publishing Company, 1967), p. 148.

 

52

two values of a random variable on a normal curve. Statistical
sampling is an advantage over any other way of choosing a part of
the population because, when the estimates of the population char-
acteristics are made from the sample results themselves, the pre-
cision of these estimates can also be gauged from the sample results
themselves.

An excellent description of all phases of sampling is included

in Sampling Techniques by Cochran.

 

Choice of a Population Sample

 

The task of determining the exact sample size is more
complicated than one would first suspect. The first difficulty con-

cerns the precision required. Ideally, the decision maker in con-

 

sultation with the researcher should decide what precision is needed
in sample size calculations. In practice this is difficult to accom-
plish: the administrator may not be able to make such a value judg-
ment due to the complexity of the problem involved, or he may not
fully understand many of the statistical implications of the experi—
ment. Based upon the experience of Shafer, 1 the recommendation

of Mendenhall and discuSSion With recreation authorities, it was

 

1Shafer, op. cit., p. 29.

2Mendenhall, op. cit., p. 158.

53

decided that sample size would be determined to ensure that
estimation error for proportions is no greater than .05 at the 95
per cent confidence level. These statistical limits were established
as the minimum acceptable, and whenever possible, the statistical
standards were set to yield more precise results.

Second, the decision on sample size will be largely governed

by the way the results are to be analyzed, so that the researcher

 

must at the outset consider, at least in broad terms, the breakdowns
to be made in the final tabulations. He can then calculate what num-
bers are needed in each subgroup to give the desired precision for
that subgroup and hence, what total sample size is required. Since
most surveys seek information on a number of different variables,
the sample size must be calculated to yield the desired information
for that "critical" question which has answers of least reliability.

The M difficulty in choosing a sample size is that in order

to make the required calculations, the variability of the population

 

must be known. Once again, sample size for the entire study will
have to be determined so as to provide an adequate number of
responses to that particular question which will have the greatest
diversity of response. Variance data is usually obtained from pre-
vious survey research among the outlined population. In the case

of this study, there was no previous research available upon which

54

to base calculations, so the maximum possible variance of any
population was used in sample size calculations. Further discus-
sion of the qualities looked for in choosing a sample can be found in

Survey Methods in Social Investigation.

 

The individual units that combine to form the sample are

referred to as "sampling units."

Usually, sampling units are speci-
fied individuals (respondents), but this does not always have to be
the case. If the sample size is large and there is not much diversity
in the number of occupants, the sampling unit can also be a vehicle.
In fact, using ”vehicles" instead of "individuals" will yield more
data for analysis. (Information can be obtained about all vehicle
occupants, not merely one person.) Because of the disadvantages
of using individuals as ”sampling units, " this study will consider
”sampling units" to be vehicles.

Although this study was designed to collect information
about the "day-user" it was administratively unfeasible to separate
"day-users" from campers at the data collection phase. Thus, the

sample frame contains both "day-users” and campers, the separa-

tion of which will take place in the analysis phase.

 

1Moser, op. cit., p. 121.

55

Choice of Sample Parks and Recreation Areas

 

Each park and recreation area can be considered to be
"individual" in that it has distinctive, unique features that attract
a separate and distinct park -using population. 1 It makes little sense
to describe an "average” user for the entire park system as the re-
sult would be a nonexistent, stereotype individual that may create
more managerial problems than were intended to be solved.

Despite the fact that parks are "individual" and should be
regarded as such, it is often necessary (usually because of insuf-
ficient research funds) to use a representative sample of all parks.
How then can the researcher choose a representative sample of all
parks in order to enjoy the advantages of sampling previously dis-
cussed? The answer is to stratify parks into similar groupings and
then choose a park from each group to represent that particular
strata. The problem arises as to how parks should be stratified.

In his survey of campers, Shafer found definite changes both in amount
of response and completeness of response from month to month dur-
ing the recreation season, and from park to park. 3 Based on his

study, Shafer recommended that parks should be stratified "on the

 

1Shafer, op. cit., p. 178.

21bid., p. 176.

 

31bid., p. 165.

 

56

basis of lake-acreage availability classes, tourist attraction densi-
ties, park-accessibility groupings and environment variability cate-
gories. "1 It has also been suggested that ”attractability indexes"

be developed and used to categorize parks. Unfortunately there have
been no suitable methods of classification developed to date, and
further research is needed in this area.

For lack of a better stratification scheme, the one proposed
by Clawson was used for this study. 2 This stratification guide,
while being so general as to not exclude any State park or recreation
area, was specific enough to indicate definite boundaries for each of
the three categories. The three classes of recreation areas proposed
by Clawson are:

1. User-oriented areas—-located close to large population
centers and characterized by a large annual attendance.

2. Intermediate areas--located within two hours driving
time from major population centers and usually characterized by the
diversity of activities offered.

3. Resource based areas—-generally located considerable
distances from large population centers and characterized by a lack

of intensive development of the site.

 

11bid., p. 5.

2Clawson, op. cit., p. 13.

 

[\9

an:

57

Upon request, lists of park characteristics and past attendance

figures for each State park and recreation area were made available

to the research staff.

A series of meetings was held with repre-

sentatives of the Division of State Parks to "pair" parks for each of

the two study phases.

The parks outlined in Table 1 and Figure 2

were chosen as both representative of the park system and "admin-

istratively feasible" enough to be included within the study.

TABLE 1. -- Sample Parks of the Michigan State Park System
Included in This Study

 

 

Class

"Experimental"
Phase

”SeasonaP'
Phase

 

1. User-oriented
area

2. Intermediate
area

3. Resource -based
area

 

Dodge Brothers
#4 State Park

Pinckney State
Recreation Area

Harrisville State
Park

 

Holland State Park

Waterloo State

Recreation Area

Tawas Point State
Park

 

Design of the Dispersal-Retrieval System

 

The limitations imposed by the Michigan State Park system

and by the use of random sampling, drastically reduced the dispersal-

retrieval options available.

The only means of dispersal available

that does not involve the use of additional staff members is by

58

SAMPLE PARKS

O "SEASONAL"
PARKS

[:1 "EXPERIMENTAL"
PARKS

teenage

@0
® 6) ALPENA.

0 ® @
TRAVERSE. .@ Q

cm® GRAYUNG 0 . HARRISV/LLE 574 r: PARK

0 @

@LUDINGTON

o e@
69 ' ® .9
s c...-

o :USKEGON

   

® 9 TAWAS POINT STATE PARK

®

.0, .GRAND @®
HOLLAND STATE PARK Q) ”Nos 0 Q) Q
_ . .
LANSING 90' 00 6‘
0

A

8R0 HERS STA TE
PARK . 4
D TROIT O

WATERLOO RECREATION AREA @m Pmcmsr RE I? .71., R ..

e @@

FIGURE 2. -- Michigan State Parks and Recreation Areas Indicating
Sample Park Locations

59

questionnaire hand out to designated sample frame vehicles at the
area entrance by the contact station ranger. Retrieval options
available are either deposit in a designated container within the
recreation area or mail-back to a central location. Since the mail-
back method is costly (at least seven cents per response), time
consuming, and requires space on the questionnaire for the mailing
address and postal permit number (thus reducing space available for
questions), this method of retrieval is unsatisfactory. The use of
follow-up waves could not be considered because there was no way
of identifying and contacting sample frame members once they had
left the park. Vehicle registrations could not be used, as previously
explained.

In summary, the questionnaire survey instrument was to be
distributed to sample frame members by contact station rangers.
Retrieval was to be by voluntary deposit in a collection box placed

near the area exit. No follow-up waves were to be used.

Questionnaire Design

 

The questionnaire should be designed so that it will obtain
the greatest response rate possible, with all answers completed and
interpretable. To accomplish this, the survey instrument must

appear to be so attractive that the respondent will want to complete

60

it, yet simple enough that all questions can be interpreted correctly
and answers can be easily given. Although adaptations were to be
made within each of the study phases, the following basic design
principles were incorporated into all questionnaires constructed.
Questionnaires were designed so as to appear brief. A
single sheet of paper stock with maximum dimensions of 8% X 11
inches was to be used. This would necessitate photo-reduction of
all printed material in order to fit within the size limitations. The
paper stock also had to be sturdy enough so that the questionnaire
could be filled out without requiring an additional supporting surface.
In addition to the information previously outlined to be in-
cluded on the questionnaire, space also had to be provided for the
particular park identification number and the date on which the ques-
tionnaire was handed out. These items were necessary in order to
identify the questionnaires for later analysis. Campers were to be
identified by their response to the question, ”Are you going to camp
in this park tonight?" Since most parks do not contain concessions
in which users can buy food staples, campers usually have to leave
and re-enter the park at least once a day. In order to cover the
situation of the camper returning to the park after buying breakfast
requirements the day he is to break camp, an additional question,

"Did you camp in this park last night?", was included. If the answer

61

to both of these questions was negative, the individual was a "day-
user.”

In addition to the traditional means of obtaining user origins
(asking the respondent to write-in his town or city, county, and State
of residence) a question requesting postal zip code numbers was
included. If respondents would answer the zip code question, a more
precise indication of origin could be made, i. e. , zip code area 48912
is a more precise area than "Lansing. "

Socio-economic level is difficult to ascertain, because it is
a composite measurement requiring evaluations from at least two
different scales. Answers to the following questions could be com-
bined to indicate socio—economic levels:

1. What is the age and sex of the "head of the family"?

2. What is the education of the "head of the family"?

3. What is the occupation of the "head of the family"?

4. What total income did the driver' 3 family earn during
last year?

Difficulty was encountered in determining which would be
the best method to obtain information about activities "participated
in" while at the park. No concensus of opinion could be found as to
which method would be best, so a pre-test was conducted with all

proposed techniques. As a result of this pre-test an activity analysis

62

question was designed in which the various activities ”participated
in" could be identified, and those activities in which the participant
spent most time could be assigned a quantitative time measurement.

As an additional aid to recreation area management, the
question "What kind of additional recreation opportunities or facili-
ties would you like to see added to the park?" was included. Also,
it was felt that this question would be an "ego booster" for the
respondent, helping to maintain interest for the activity analysis
question that followed.

In order to provide background information for later market
studies, Michigan State Park authorities indicated that they wished
to obtain guidance about why people actually chose a particular park.
The question, ”Why did you choose this park rather than a different
one?" was included as a means of outlining answer categories to be
used in future studies.

Throughout the entire survey, a central motivating theme
and symbol were used. The symbol was a picnic table silhouette
which was superimposed on all questionnaires, signs and collection
boxes. It was hoped that respondents would subconsciously associate
the questionnaire to the instructions on the signs and ultimately to
the collection box, thus improving response. The theme, "by com-

pleting and handing in this card you can help plan your parks" was

63

expressed in the questionnaire, in signs, and in the short verbal

message given by the contact station ranger.

CHAPTER IV

ACCURACY OF COLLECTED INFORMATION

The purpose of this study phase was to test sub-hypothesis
two, that is, to check the accuracy of information collected by ques-
tionnaire. In order to accomplish this, a continuous data collection
system using a questionnaire survey instrument was designed and
implemented. The data collected from this system was then com-
pared to parameter data for the park-using population. Thus, the

accuracy limits of questionnaire information were determined.

Preliminary Preparations

 

Que stionnai re De sign

 

The requirements of the ”seasonal" questionnaire were that
it must be attractive, distinctive, convenient to complete, and diffi-
cult to forget or ignore. Respondents could not fill out the question—
naire immediately upon receiving it at the area entrance because of
the activity analysis questions. Motivational problems would be

created because the respondent could not fill it out immediately while

64

65

interest was high; rather it had to be set aside until all activities had
been completed. In order to overcome this difficulty and the ones
previously outlined in Chapter III, a number of devices were employed.
m, the questionnaire was printed on bright yellow 24 pound Carni-
val Kraft stock. This was a bright, distinctive, relatively inexpen-
sive paper stock with enough rigidity so that the questionnaire card
could be completed without requiring an additional supporting surface
during the writing. Second, the card was a single sheet of paper
stock folded once to provide a brochure-like questionnaire with dimen-
sions 5% X 8% inches. Thus, the questionnaire appeared to require
little time for completion and the folded size required less supporting
surface during completion. m, a slotted hole was drilled in the
upper left-hand corner of the card, permitting it to be suspended from
a vehicle instrument knob. This would aid response by keeping the
questionnaire in view of the vehicle occupants, thus reducing forget-
fulness and misplacement. Fourth, a golf pencil was supplied to re-
spondents who indicated that they needed a writing instrument.

A sample of the "seasonal" questionnaire is contained in

Appendix A.

Sample Siz e

 

Since information as to the variability of the Michigan State

Park and Recreation Area "day-user” was not available, sample size

66

. 1 . .
was calculated uSIng absolute error and assuming the max1mum

possible population variance. Calculation was as follows:
The probability requirement

Pr[1p-p1\<.05] 2.95

leads to
lp-pl .05
P13 /p (1 -p) Sq /p (1 -p) 2 .95
n n
Where

p = unknown proportion to be estimated

estimate of p

’U)
H

n = sample size
.05 = absolute error

.95 = confidence level
Using normal approximations we have that n must satisfy

05

.. /p(1-p) 21.96
11

 

1As contrasted to relative error where the accuracy is
specified in terms such as .05p and where lower bounds for p are
needed in order to determine sample size.

67

fr? >/ m (39.2)
n >/ 1538 p (1 -p)

Since p(1-p)\<%

Then n 3%,- - 1538 = 385

For each unit of analysis, 385 "day-use" respondents were
required. In this study, each park and recreation area was a separate
analysis unit. Also, the occupants of each vehicle using the area
entrance were not necessarily "day-users" (they might have been
campers or park staff). Therefore, a sample of at least 900 respon-
dents was set for each park and recreation area to guarantee the

necessary ”day—use" sample size.

Sampling Method

 

Systematic random sampling was chosen as the method to
be used in this study phase because it could be implemented with
least interference to normal recreation area operation. In system-
atic sampling, every nth member of the population was included in
the sample. For instance, to obtain a sample of 250 vehicles from
a population of 1000, every fourth vehicle was selected. To ensure

that every vehicle in the population had equal opportunity of being

68

chosen, the first sample vehicle had to be designated through the use
of a random choice method (1. e. , random number tables, "drawing"
from the sample space). After the initial vehicle had been identified,
every fourth vehicle thereafter would automatically be included in the
sample.

Since contact station rangers do not have the time to count
all vehicles entering the area, a simple means of counting vehicles
and identifying sample members was designed. A pneumatic traffic
counter was adapted, through the use of slot car gears, to count
axles and then to ring a bell when the predetermined number of axles
was reached.

To determine the setting of a traffic counter, the following
procedure was used: For each recreation area, the attendance for
each summer month during the past three years was obtained. From
this, a "conservative" estimated attendance for the months of July
and August, 1968, was set. The required sample size was then
divided into the estimated attendance to provide the systematic sam-
pling frequency. Through experimentation with numerous gear ratios
in the traffic counter, a setting for each individual park was obtained.
For example, Holland State Park required a systematic sample of
every fifty-fifth vehicle. A gear ratio to provide this exact sampling

frequency could not be constructed, so one was constructed which

69

would yield a greater sample size than was actually required. In
the case of Holland State Park the traffic counter identified each
one hundred and fourth axle (fifty-second vehicle).

Some vehicles entering recreation areas have more than
two axles, i. e. , card with boats, trailers or campers. Since
vehicles with more than two axles apparently do not enter recreation
areas in a definite order or scheme of stratification, they do not
jeopardize the random sampling system. 1 In fact, they increase
the number of respondents surveyed, thus helping to guarantee the
filling of the sampling frame. If the questionnaire was distributed
to the vehicle that caused the traffic counter bell to ring, those
vehicles having more than two axles would have an advantage. To
overcome this problem, the second vehicle following the one that
rang the bell was designated to receive the questionnaire.

Implementation of the Continuous
Data Collection Sgyglem

 

 

One month before the system was scheduled to be put into
effect, a meeting was held with all sample park supervisors, regional
park supervisors, and representatives of the Recreation Resource

Planning Division and State Park Division of the Michigan Department

 

1Cochran, op. cit., p. 230.

70

of Natural Resources. At this time, the study objectives were
outlined, possible operational problems were discussed and solutions
were proposed. Special emphasis was given to possible implications
of the collected data for future recreation area management.

Using the discussion from the meeting as a guide, written
instructions were prepared for use within each sample park. These
instructions were constantly updated as new problems were encoun-
tered. A weekly report form, included in Appendix B, was also
designed. This form, in combination with the regular weekly report
forwarded to the State Park Divisional Office, provided the following
information on a daily basis: number of visitors and vehicles enter-

" number of

ing the park, number of campers, and "day-users,
people refused entrance to the park (usually because of a capacity
crowd), number of questionnaires handed out, number of question-
naires returned, the presence of any special events in the region of
the park that might influence attendance, and the weather—-including
amount of cloud cover, amount of precipitation, amount of wind, and
maximum / minimum temperatures.

Each sample park was visited by a study team member
shortly before commencement of the data collection system. The

park supervisors were asked to delegate one park staff member as

the "contact man" for the study, and permission was obtained to work

71

directly with the delegated person. In most cases, the contact man
was either the park supervisor or assistant supervisor. The con-
tact person was given the responsibility for ensuring that all phases
of the study were carried out correctly. If any problems arose that
were not outlined in the written instructions or orientation discus-
sions, study team members were to be immediately contacted by tele-
phone.

During the initial visit, study team members supervised
the installation of the traffic counter, signs and collection box. The
traffic counter was placed in the contact station and set with the pre-
viously determined random starting position. A short distance past
the contact station a sign was erected that asked respondents to
complete the questionnaire and return it at the area exit. It was
anticipated that as one occupant of the vehicle was accepting the
questionnaire, the other occupants could read this motivating Sign.
The ”deposit" box was placed near the park exit, usually on the pas-
senger side of the roadway. A short distance in advance of the
deposit box a sign, ”Please Deposit Completed Park Use Card
Ahead, " was posted. At the collection box an additional sign, "Please
Deposit Park Use Card Here, ” was erected. The tasks of the contact
station rangers were demonstrated: the placing of park code num-

bers and hand-out dates on each questionnaire to be done in the

72

 

FIGURE 3. -- Vehicles Waiting to Enter Dodge
Brothers #4 State Park (A Typical
Weekend Scene)

 

FIGURE 4. -- Contact Station Rangers Handing
Out Questionnaires (Standing Out-
side the Contact Station During
Heavy Traffic Flow)

73

 

FIGURE 5. -- Handing Out Questionnaires While
Remaining Within the Contact
Station (Flexibility Is An Asset)

 

FIGURE 6. —— View of Motivational Sign and
Pneumatic Traffic Counter Hose
From the Contact Station

74

   
    
 
 
  

  

-- ' ' .
..‘u - .,T o ' .4 " ' i n 1‘ “_ 1“

.- “L

‘ Tit ‘ We! mom comma

   

PARK-USE CARD

AHEA'D

 

FIGURE 7. -- Sign Warning Motorists of Their
Approach to the Deposit Box (As
Motorist Is Leaving the Park)

PLEASE DEPOSIT COMPLETED
PARK-USE CARD

     

 

FIGURE 8. --Deposit Box and Sign (Located
Near Park Exit)

75

 

FIGURE 9. -- Pneumatic Traffic Counter
Modified for Systematic Sampling
Use (Designed So As to Ring a
Bell Every nfl Axle)

. ’. 1 .
g ‘ - ._ c
. “- ‘ t.{'t

’ --
- 7

-'
.;.¢

’

 

FIGURE 10. --Workers at the Study Center
Preparing Experimental Ques-
tionnaires for Shipment to Parks

76

morning before the vehicular traffic became busy; the completion of
all forms; the collecting, packaging, and mailing of returned ques-
tionnaires; and the motivating message to be given when the ques-
tionnaire was handed out was rehearsed, namely, "Good day, Sir.
Would you please fill out this card and leave it in the box near the
exit as you leave the park? It is important that you complete every
question. Do you need a pencil? Thanks for your cooperation—-
enjoy your visit. "

Sample parks were re-visited by study team members

throughout the duration of the study.

Collection of Parameter Information

 

Parameters of the park-using population were to be obtained
by a combination voluntary "hand in"-interview during the latter part
of August. Questionnaires were to be handed to all vehicles entering
the park during the interview period. A "spotter” was to be located
within viewing distance of the collection box and in a position to sig-
nal to the interview team farther down the road. Those "vehicles"
voluntarily returning questionnaires in the deposit box at the park
exit were allowed to leave the park. The ”vehicles" not depositing
questionnaires in the collection box were to be stopped and inter-
viewed to obtain information for certain "key" questions. The com-

bination of data collection methods outlined above should produce

77

near 100 per cent response from all recreation area users during

the interview period (hereafter, termed "parameter" information).

Implementation of Parameter
Information Collection

 

 

Supervisors of the "seasonal" parks were contacted two
weeks before the parameter information was scheduled to be collected,
and assistance, in the form of making park staff members available
to aid in the interview program, was requested. It was proposed
that the interview teams consist of five members: one person (pref-
erably a park staff member) would hand out questionnaires to all
vehicles entering the park; one person would be positioned about 50

"spotter" (again, preferably a

yards past the deposit box to act as
park staff member). Those vehicles not depositing questionnaires
would be directed into an interview area; two persons would be con-
ducting interviews; and one person would be supervising the opera-
tion and relieving other team members.

During the first interview day it became apparent that the
outlined interview program was inadequate. Traffic flow at the area
exit was erratic, requiring five or six interviewers at times. In

order to accommodate the sporadic vehicular flow, a supervised

completion of a "short-form questionnaire" was substituted for the

78

verbal interview. 1 Those ”vehicles" not depositing questionnaires
in the collection box were directed into a designated area. The
driver of the vehicle was then asked to complete the "short-form
questionnaire. ” The questionnaire was collected when completed,
the driver was thanked, and the vehicle was allowed to proceed. The
success of this method led to one additional change--another collec-
tion box was erected at the exit of the designated interview area and
drivers were asked to deposit the "short—form questionnaire" when
it was completed.

At Holland State Park the intense vehicular traffic neces-
sitated a further change. Even with three people handing question-
naires to vehicles entering the park, there was a traffic line waiting
to enter extending for one-half mile. Experimentation revealed that
there was no possible means of dispersing questionnaires at the area
entrance without creating a major traffic congestion. The problem
at the park exit was similar to that at the entrance--too much traffic
for the one-lane exit. Therefore, it was decided to use random
number tables to choose a random sample at the area exit. Those

"vehicles" included in the sample were directed into a designated

 

1A sample "short—form questionnaire" is contained in
Appendix C.

79

area, asked to complete a questionnaire, and to deposit it in the

collection box provided at the area exit.

Treatment of Collected Data

 

Collected data was to be placed on data processing punch
cards for the reasons previously outlined. Two methods could be

employed to accomplish this task. The first method is through key

 

punch-—questionnaires have to be precoded by study team members,
coded key punch tables must then be prepared, and finally, key
punch operators use the prepared coded tables to punch the data
processing cards. With this method there are three sources of
human error-~the preceding, the preparation of coded tables, and
the punching of the data processing cards. The second available

method is through the use of optical scan sheets—-questionnaires

 

must still be precoded by study team members, but instead of plac-
ing the coded information in key punch tables it is put on I. B. M.
optical scan sheets. The optical scan sheets can then be processed
directly through a reading-card punch machine to produce the desired
data processing cards. Using this method there are only two sources
of human error (the key punch operator is eliminated), and because
there are fewer human handling procedures the cost is less than the
key punch method. These advantages were sufficient reason to use

the optical scan method of data treatment in this study.

80

Special I. B. M. optical scan forms were designed and are

included in Appendix D of this report.

Analysis of Data

 

Only those questionnaires which were entirely completed
were coded and included in the analysis. If one question was not
filled in, and it was not among those questions judged to be essential
for analysis (i. e. , questions 1, 2, 4, 6, 7, 8, 9, 10, 11, 12, 13,

14, 21, 23, 24 were considered "essential"), the non—completion was
regarded as an error and the questionnaire was coded.

Parameter information was compared to the information
collected by voluntary deposit of questionnaires by means of analysis
of variance and chi square techniques. The responses from ques-
tions possessing "continuous" answer categories were analyzed
through the use of analysis of variance. The chi square technique
was employed on the remaining data.

Information from each sample "seasonal" park was analyzed
in two different time periods:

1. Data collected from voluntary deposit1 throughout the

entire testing season were compared to the parameter data.

 

1Data collected from the dispersal-retrieval system pre-
viously outlined in the chapter is hereafter referred to as "voluntary
deposit data. " Voluntary deposit data, therefore, consists of the
completed questionnaires that were voluntarily deposited in the

81

2. Data collected from voluntary deposit during the week
immediately prior to, and the week immediately following parameter
information collection, were compared to the parameter information.

In order to test the accuracy of the voluntary information,
only the second time period comparison is needed. The first time
period comparison was only intended to test the changes in use and
user characteristics which occur throughout the duration of the sum-
mer season.

Within each of these time periods the "day-users" were

analyzed separately from the campers.

Results and Conclusions

 

The random-systematic sampling procedure used in this
study phase appeared to function successfully. Minor difficulties
occurred but were solved with a minimum of trouble. Due to the
difficulty involved in stopping sample vehicles as they enter the park
(vehicles with season passes do not have to stop at the park entrance),

from 3% to 15% of the sample vehicles were missed. Generally, this

 

collection box near each ”seasonal" sample park exit (with the excep-
tion of the data obtained during the few days that parameter informa-
tion collection took place).

1At Holland State Park, the "voluntary deposit" system was
not re-activated when the interview team completed the collection of
parameter data. Data collection from voluntary deposit within the
two weeks immediately prior to the parameter information collection
were used in the comparison analysis.

82

proportion was under 5%, but as the amount of traffic increased, one
might suspect that the number of ”missed" sample vehicles also
increased.

A summary of the number of questionnaires distributed and
returned during the "seasonal” testing program is contained in
Tables 2 and 3. It appears that there is an inverse relationship
between the number of completed, returned questionnaires and the
total number of people visiting the park. This may be due to the fact
that unlike the low-use parks, the staff in the high-use parks did not
have the time to deliver the required motivating message. Also, a
small sample included in a large population tends to lose the feeling

' There is less risk of "social dis-

of "supervised responsibility.‘
approval" resulting from not handing in the questionnaire (it is
easier to lose oneself in the crowd).

Parameter information collection was conducted on the dates
outlined in Table 3. Since financial resources were limited and the
amount of information that could be obtained at Tawas Point State
Park was small, it was decided to curtail parameter information
collection at that park and to move the interview team to Holland
State Park. Shortly after the arrival of the team at Holland State
Park, the weather became cold and windy, and the number of visitors

declined drastically. However, over a period of five days, enough

data were collected so that the desired analysis could be conducted.

83

.30 popes: mofimnsofimmav mo son
-85: mg 0“ moawmcsoimosc pouoafioo Mo .8285: mg maimmfioo L3 poswwuno 93mm mum.

 

 

 

A

3mm 53 SS 38. 3.8a 683860

can 3 was .3 .mm :3 $3 spam 83m eemzom

so .5. S: s 3 Sam 33 spam 9.6% EBA page

can .3 ma .3. .3 mg 32 mpp<6mmwmwmmmeB
emwmfim emwmmmmwsa ”flaw emwwmwé ewmqwfiwe is

H

 

mo ponfisz

 

 

mo ponfizz

 

mo sonasz

 

 

 

AwuwQ aowoamsmm wcfipfioxmv mxpwm onEmm :Hmsommom:
555». 339800 powpsh paw .posaaom J50 @0933 moawmacoflmosmu mo amnESZ I .m gamma.

84

TABLE 3. -- Number of Completed Questionnaires Obtained from
”Seasonal" Sample Park Users During Parameter
Information Collection

 

 

 

 

 

 

 

 

 

 

Park Date Voluntary Superv1sed
DepOSIt Response
Waterloo State Recreation Area August 18 231 70
August 19 141 37
TOTAL 372 107
Tawas Point State Park August 21 51 6
Holland State Park August 24 8 204
August 25 75
August 26 43
August 27 73
August 28 52
TOTAL 8 447
COMBINED PARKS TOTAL 431 560

 

 

 

85

Analysis of the collected data revealed that each park has
user characteristics that are "individual" and characteristic of that
park alone. It was also evident that even within each park, the

characteristics change throughout the season. The visitation time

 

during the season was shown to be closely related to the type of per-
son who used the parks and the way in which the park was used.
The accuracy of information collected by voluntary deposit

"voluntary” data to the collected

was determined by comparing the
parameter information. If the assumption is made that the parame-
ter information is valid, then any statistical difference between the
two sets of data would indicate that the "voluntary" data was not
accurate. However, because of the circumstances involved during
the collection of the parameter information, a great deal of caution
must be exercised in analytical interpretation.

Information contained in the questionnaire yielded sixty-
three different variables for comparative analysis. The results of
this analysis indicated that the accuracy of ”voluntary" information
was high. Even with the comparatively low amount of return, and
with no "follow-up" waves to reduce non—response, the comparison
analysis showed very little difference between ”voluntary" data and
parameter data. At the 95% confidence level there was no significant

difference between the two groups except for the few variables out-

lined in Table 4.

86

TABLE 4. -- Variables from "Day-User" Questionnaire Information
Which Show a Significant Difference in a Comparative
Analysis (From Use of Both Analysis of Variance and
Chi Square Techniques)

 

 

Park Variables

 

Tawas Point State Park None
(low-intensity use)

 

Waterloo State Recreation Area 1. How many miles, by the
(intermediate-intensity use) most direct route, is this
park from your home?

 

Holland State Park 1. How many days have you
(high-intensity use) used this park in 1968?

2. How much time did you
spend traveling to this
park today?

 

 

87

The subject matter area in which the voluntary deposit of
questionnaires yielded the least accurate information about ”day—
users" was in the characteristics of travel to the park. The "sea-
sonal" questionnaire contained two questions requesting travel
information: "How many miles, by the most direct route, is this
park from your home?" and "How much time did you Spend travel-
ing to this park today?" In data obtained at Holland State Park the
"travel time" question proved to be significantly different in the com-
parison of ”parameter" and "voluntary" information. The ”miles
traveled" question showed a significant difference in information
collected at Waterloo State Recreation Area. Surprisingly, at both
of these parks only one of the two travel questions indicated a sig-
nificant difference. The other travel question, in the case of both
parks, provided responses which showed close agreement.

There are two possible causes for the lack of accuracy in
the "travel" questions. If the park-using population does not know
(or care about) the distance traveled and time taken to reach the
park, then the information they provide will not be reliable. Perhaps,
when people are engaged in "day-use" activities they are more con-
cerned with enjoyment than they are with travel measurements. The
second possibility is that the population receiving interviews is more
methodical (or accurate) in arriving at answers to the "travel" ques-

tions. An inspection of analysis results shows that the average

88

information obtained by interview indicates less ”distance traveled"
and ”time taken" than information obtained by voluntary deposit.
The presence of the interviewer appears to ”trim down” measure-
ments of the distance traveled.

Since there was no "voluntary deposit" information available

for Holland State Park after the collection of parameter information,

 

one would expect the variable "How many days have you used this
park in 1968?" to Show a lack of accuracy. In the other "seasonal"
sample parks, the "voluntary deposit" information was "balanced"
around the parameter information. This was accomplished by in-
cluding in the comparison analysis "voluntary information" collected
prior to, and immediately after parameter information collection.

H H

It was not possible to "balance voluntary data" at Holland State
Park since no voluntary collection of data took place following parame-
ter information collection. If the park using population consists of
individuals who use the park often, an "unbalanced” time comparison
would show a significant difference in response to this question. An
examination of the data revealed that this, indeed, was the case.
Parameter information showed a slight increase in the number of

days the park was used when compared to ”voluntary deposit" data.

This increase was enough to indicate a significant difference between

the two sets of data.

89

It appeared that as the intensity of usage within a particular
park increased, there was a decrease in the accuracy of information
collected by means of the questionnaire system outlined. It was also
noted that in all parks and in all time period comparisons, more
variables within the "camper" group were significantly different than
among the "day-user" group. This indicates that accurate "day-
user” information can be obtained from a smaller sample size than
can accurate information about campers. Therefore, when continu-
ous data collection systems are being designed which will collect
information from both campers and "day—users, " the sample size
should be determined for camper information.

The inclusion of a message asking respondents to mail the
questionnaire back to a central point, if they have forgotten to deposit
it while in the park, appears to increase response by less than 1%.

Sub-hypothesis two, as previously stated, reads:

Questionnaires returned within the proposed dispersal-

retrieval system will contain information that does not differ
significantly from the parameters of the park -using population
of the Michigan State Park System.

From the results of this study phase, it appears that this

sub-hypothesis cannot be rejected.

Repetitive Entries. -- The phenomenon of ”repetitive entries"

 

was of more importance than had first been suspected. This practice

90

was more prevalent among campers than among "day-users. " The
lack of a grocery store within a state park resulted in many campers
having to leave the park in order to obtain food supplies. During the
period when interview teams were in the parks it was noted that some
vehicles left and re-entered the park more than six times per day.
Indeed, within some state parks, the traffic on the main thoroughfare
resembled a parade. Cars full of youngsters continuously cruised in
and out of the park searching for some "action. "

The sale and use of "season passes" for entry into parks
encouraged the practice of leaving and re—entering. Since it was

"repetitive-entry" vehicle during sur-

impractical to determine the
vey work, the results of all traffic-type surveys will over-represent
the ”repetitive-entry" population. The extent of this bias will depend
upon the relative size of the ”repetitive-entry" population, and the

difference in characteristics between the general park-using popula-

tion and the ”repetitive—entry" population.

CHAPTER V

DETERMINATION OF "BEST" QUESTIONNAIRE

AND DISPERSAL-RE TRIEVAL SYSTEM

The purpose of this study phase was to attempt to overcome
the main problem of the questionnaire survey method--the vital prob-
lem of non—response. To accomplish this end, those variables which
might affect the number of questionnaires returned were identified.
The variables were then tested, one at a time, in order to measure
the effect of each on response rates.

Determination of Variables Which Might Influence
Questionnaire Response

 

 

The many factors which combine to influence questionnaire
response rate are complex and difficult to isolate for study. As in
all research relating to motivations, the examination of these factors
is still in a very elementary stage. Little is known other than the most
elementary relationships involved, namely, response depends upon
the population receiving the questionnaire, the appearance and content

of the questionnaire itself, and the circumstances in which the

91

92

questionnaire is received and collected. These factors have been
discussed in detail in Chapter II.

In order to ascertain the variables that should be tested in
this study, a review of factors that had influenced questionnaire
response in the past was conducted. Also, a meeting with park and
recreation authorities provided information on further factors which
might affect amount of return. From these two sources, the vari-
ables in Table 5 were outlined as being capable of being tested in this
study.

Means of Choosing "Best" Alternatives
Within Each Variable

 

 

Within the context of this study phase, "best" is interpreted
to mean that means by which the greatest number of completed ques—
tionnaire returns were obtained. To determine the "best" means of
minimizing non—response, the alternatives within each variable being
tested were outlined and questionnaires were designed utilizing each
alternative. For example, if color of paper stock was the variable,
the specific colors to be tested (1. e. , blue, white, yellow, etc.) were
the alternatives within that variable. The other test variables which
might affect response were held constant. Therefore, if the ques-
tionnaires containing the alternatives of one variable were handed

out in a random manner, one could state that the alternative which

93

 

 

 

 

 

 

.3388sz o: Bog 66:on oz E
.m>
mmofisfiowpfi so $3 8. one
omcoamoanco: wswospop hammom 3:: Jenna oawmscofimmsv mo
“sofiapmfi mg 89% mafiwccoflmmsv 9: economy. sosaoo tea soda: 5 Bo: pmfiofim 3
Cu acopcommop o5 poBOZw 30: pofiofim m
sowam £293.“ 16.52“me UGJOQ em £3 pcm owcwao xoon Roman 90 p300
550.5 mafia 303mm 353 m0 somEmQEoo
xooum ao>oo came x285 zoopm pamo :bwwom: AZ
.m>
spawn co m o m
Messages? poaofioo «HEM Hmzsawv eased em x p cum :3 5: 3
86:6. :2; 2 x mm 6628:: p.28 36E E
.m>

 

“3235.3
:65 we x we 6 9S 6629 562.... :2: A A x mm
a

 

28 ppEom

3

 

Aooswswoaag Ransom ofiwcsoflmosa

 

mo>flmcaofi< .Ho soflaflomom

 

£0.ng

 

 

.8 63860 6.83 SEE ppm mmcoAmpm 855836650 Spas 232 533 SHARES, -- .m Humans

mpgm mEH E @3on msfiom

94

 

.ommmmoa :fimoaop momma: m magnum”?
1:00 .566 aofio 9.3 690. came pogo.“ on“ so
msoflmosv of .Ho poapmsvnoso bouwgwxoaam mi

 

 

usfimusoo can one 8:? came “won :05 m X m m mafimcsofimosv on? 26.30 Rom: 3w
.m>
came :05 3 X mm poEoM enamosoflmosfio came poEob 3 .m
msofimosv ofimm o5 .
sopso sofimosc uneaomfim 0

mo mmsflopao “Genomic moat mo someano

 

msofimosv mfimbwsm 533“;
m0 mayo.“ “Genomflp oz.“ mo nomfiawmaoo

 

soflmosv mwmbwsm
33306. am we 298m «coaomflm um

 

“sousou ohmccofim mama

 

©3395 Zosom 02 A2

.m>

 

Ugandan Sodom .200

oawmgofimosv fits poflamsm Sodom Q .m

 

mo>3mcsofi< mo :oEQEomoQ

 

6333,

 

 

UmsnficoU I. .m Himmdwh.

 

 

 

 

ARE .5 ma
Eugen mo msmoa Eco mg. was @95an ohms sages
meson. fimoaom .mfixo mops pm paw mocmnpso .
. . sown Ewe fixo mops sum psopcmm At
am “so popcmm oEm pommoappwuflom Ema
Iona fits 09$ :Pfimo amen: mm? mummacoflmosmu
.m>
.owmmmoE
:fimomop ommoa: fits on? :ppmo “mom: was “the “a son. sotoofioo 5
opwwgoflmosg .mpgm mo omega :Hmcommom: fimoaop .oocwauco moss am #50983 3 A.
5 com: Hmsoflwesufimmpoamfip mo “80532
.sofimoofi pommoappmuoam one op
% cocowcossoo m3 um peso one. .288 3500 so mops sown 2mg .8 tone mesa
ofi $3 on ma xon cotbozoo 5 «Somme sonic pm xon sotoofioo a: fimomop Ann
3500 Sampsoamom .oEm pmmmosppmufiom .Emm Hm>oflpoa .oocmfism «m «noesmm At
no.3 fits on? :ppmo Eon: was oawmqaoflmosg
.ms
.owmmmoS
:fimoaop emmoaz fits on? :pawo Eon: was “the mesa Hm Non coflooHHoo
efimccoflmosg .mp3»). mo ommnm :Hmsommom: E fimoaop .oosmfiso “m psopcmm 3 .H
5 pom: Hmsoflpopéwmaommwp Ho @0532

 

 

mpofioz ~m>oEuom1HmmsemmflQ

 

mo>fimssofi< .Ho cothrnomem o3m€m>

 

 

 

possthou I .m gamma.

96

achieved the highest response rate could be said to be ”best. " A
description of all ”experimental" questionnaires is contained in
Appendix F.

In order to ensure that the results were due to the specific
alternative tested and not due to chance, a "reasonably large number"
of questionnaires were handed out and statistical techniques of sig-
nificant differences were employed. The method used to determine

whether one particular alternative was actually "better" than another

was the ”fixed-pair" method of statistical significance:

X.
Pi = Kl- = proportion of the number of type i question-
i naires that were completed and returned
x = number of completed, returned question-

naires of type i

N = number of type i questionnaires handed out
When comparing Pi and Pj , Pi - P3. is approximately normally dis-

2 2 2
tributed1 with mean Pi - Pj and variance / ij = &/ i + a/ j , where

Pi(1 - P.) P.(1 - P.)

A 2
[21: N. 1 is the variance of Pi and/ j = L N. J is the

1 J

 

 

 

1For proportions of the magnitude that will be estimated,
30 or more trials will ensure approximate normality.

97

variance of Pj . Using the fact that P(1 - P) \< :1— for all 0 \< P \< 1,

we can conclude that é/ij \< —— + —— .

 

A 95% confidence interval for P1 - Pj is Pi - Pj - 1. 96$sz + —L ,

 

* C 1 1 A . . . . .
Pi - Pj + 1. 96\/4:I\Ti + 4N; and Pi is Significantly different from

PJ. if this confidence interval fails to include 0. For example, with

N = N. = 500, the 95% confidence interval is P. - P. - —1——Q§

1 J 1 3 V 1000
A A 1.96 A A A A
P-P+—— or Pi-P.-.062,Pi-Pj+.062 ,so

i j V1000 3

there is a significant difference at the 95% confidence level if
113. - 13.1 > .062.
1 I

It is known that the answer given to a question depends upon
the respondent' s interpretation of the question and his ability to
answer. Presenting the same question through different methods of
data collection may result in different answers, depending upon the
interpretation of the question in each context. There is a possibility
that the population replying to a particular question on a blue ques-
tionnaire may differ from those who answer the same question on a
white questionnaire, even if the only difference between the two ques-

tionnaires is the color of the paper stock. This variability in response

98

can be reduced by designing questions so as to be simple to answer

and unambiguous.

Sampling Method

 

The sampling method used had to provide a sample that was
representative of the park—using population. For comparison pur-
poses, enough questionnaires had to be distributed so that the returns
would not result in a bias due to choosing a small, select population
as the sample. This problem was eliminated by printing and distrib-
uting to the sample parks at least five hundred copies of each alter-
native questionnaire. Since the different questionnaires had to be
distributed in a random manner, they were collated together before
shipment to the parks. The contact station ranger then handed out
the questionnaires in order, beginning at the top of the collated stack.

In an attempt to ensure that the sampling period provided
results which were representative of the total park-using population,
all tests were scheduled to include at least one weekday and one day
of the weekend. Of the total questionnaires to be distributed in the
test, 80 per cent were handed out on the weekend day and 20 per cent
were handed out on the weekday. These percentages reflected the
proportion of people visiting sample parks on weekdays and weekend

days.

99

A further use of proportionate distribution was made in
determining the exact number of collated questionnaires to be sup-
plied to each park. A survey of past attendance records for the
sample parks revealed that the total number of questionnaires in
the test should be distributed as follows: Dodge Brothers #4 State
Park--60%, Pinckney State Recreation Area--30%, Harrissville State
Park——10%.

The sampling periods during which questionnaires were to
be handed out were selected as follows: For each test, the sampling
days were divided into morning (8:00 a. m. to noon) and afternoon
(noon to 10:00 p. m. ). The number of visitors entering parks during
these time divisions was approximately equal. A random selection
of the hours within each division was used to obtain the starting time
for each sampling period. Questionnaires were then handed out until
the supply was exhausted. In the case of inclement weather, sampling
periods were adjusted so that questionnaire hand—out began as soon
as the contact station was open.

One questionnaire was handed to the occupants of each
vehicle entering the area during a sampling period. There were two
exceptions to this procedure. Due to the small number of question-
naires to be distributed in Test Seven, every second "vehicle” enter-

ing the area received a questionnaire during the sampling period.

100

The investigation of mail back return, which took place in Test Eight,
was conducted with questionnaire distribution at the area exit. Dur—
ing sample periods a park ranger was stationed at the area exit and
handed prepaid, self—addressed questionnaires to all vehicles leav—
ing the recreation area. Thus, the only means of returning the ques-

tionnaire was by mail.

Implementation of the Testing Schedule

 

In order to complete the testing of all variables within the
summer months, an intensive testing schedule was outlined. This
schedule is contained in Table 6.

As in the "seasonal” study phase, all park supervisors from
the "experimental" parks attended an orientation meeting prior to the
commencement of testing. Sampling and testing techniques were dis-
cussed, and the importance of close cooperation was stressed.

Study team members visited the ”experimental" sample
parks to supervise the installation of signs and collection boxes.
Placement of these boxes and signs was identical to that used in the
"seasonal" study phase. A ”contact" man was chosen, and since
this person had to assume a great deal of responsibility, much time
was spent outlining sampling, testing and analysis procedures. The

"contact" man was instructed to immediately contact study team

101

 

30: o: .m> 30: pofiOHm

Axomoflflmfi
pew 5-955 “Swag peso umom .m> fimsofi ._.H.m3mom__

Antenna 8.32 P86 Hop .9» £32 in???

S3 .3 DE. $882

33 .mm .33. seesaw

 

xooum .8an mo soHoU

$2 .5. .23. 5623.8
$2 .3. b3. seeps

 

pomofioeo 30qu o: .m> Zosem

33 .NN ER. $882
$3 .AN 33. $856

 

msofimosv mwmbmsm btfipow “soaomfim

$3 .3 .33. .smppsemm
$3 .2 has .smetm

 

ofion o: .m> 30: @39on

xoopm peso :bsmosz .m> xoopm Paco :Ewflq:

psmo pogomcs .m> pogoh

$2 .3 .23. seesaw
$2 .2 has .smepfimm

82 .2 .33. 582m

 

Umpm OH. mOHQmwhm>

 

moped

 

Emu.

 

 

Emma .gflemfitmmxmz so 6388..“ I .o mamas.

 

m3: .3 “mums... $8..me
53ng 512393 .m> champ 333.23% I h. wmmfi .m 55.954 Swank w

 

102

.H. @3339» 5 303.3938 yo.“ mama m mm. Umms mpm?
335mg 23. .mfiwccofimwsw :HmGOmmmm: 8. mmcommmp 90 5mm. 33 .3. $395. .hqunm N.

 

32 .m “mama... .3233
mpmvuo c3525 Emummfim . O wmmfi .N pmfiwfiw Swank o

 

Umummh. mmfinwflmxw mmme «mmH

 

 

 

 

vmscficoU nu .w madmgfi.

103

members by telephone if there was any doubt about testing procedures.
To ensure that the data collection system was functioning correctly,
study team members were present in all sample parks during the
first "experimental" test. Throughout the remaining tests unan—
nounced visits were made to sample parks in order to assess the
success of testing procedures.

The weekly administrative procedure during the series of
experiments was as follows: Questionnaires utilizing the alternatives
to be tested on the following weekend were designed on Sunday. On
Monday, the ”photo—ready" questionnaire designs were delivered to
the printer who generally had the desired number of copies prepared
before Tuesday afternoon. The questionnaires were then rushed to
the study center where park identification numbers and the date were
stamped on them. Following collation of the different questionnaires
within each test, they were counted, sorted and prepared for ship-
ment. A set of instructions outlining sampling procedures, question—
naire identification, and the recording of results were enclosed with-
in each shipment package. A set of typical park instructions is
enclosed in Appendix E. The "experimental" sample parks received
the packaged questionnaires on Thursday afternoon in preparation
for hand-out during the testing periods. At the completion of each
test, the returned questionnaires were forwarded by the park staff to

the study center for analysis.

104

Analysis of Data

 

As questionnaires were returned to the study center they
were examined for completeness of response. 1 Using the statistical
methods previously outlined in this chapter, the significance of each
variable was calculated.

In order to determine whether the population completing one
particular type of questionnaire differed from the population complet-
ing another type of questionnaire (within the same experimental test),
a comparative analysis was conducted. The responses to the five dif-
ferent questionnaires in experimental Test 2 were compared using
the analysis of variance and chi square techniques previously
described in Chapter IV. The comparison would reveal differences
in response between the different questionnaire types and would also
reveal differences in response among the different "experimental"

sample parks.

 

1The criteria used to judge completeness of response was
the same criteria used in the ”seasonal" study phase. In Tests 1,
3, and 4 all questionnaires which were returned were judged to be
complete. In these particular tests, the variables being investi-
gated did not affect the actual content of the questionnaires. There-
fore, there was no reason to suspect that there would be a difference
between proportion of questionnaires returned and proportion of ques-
tionnaires completed.

105

Results and Conclusions

 

All of the experimental tests were conducted as scheduled,
and no major difficulties occurred during the testing program.

A summary of the number of questionnaires that were handed
out, completed and returned is included in Appendix G. It is apparent
that, similar to the results of the "seasonal” study phase, as the num-
ber of people visiting the park increases the proportion of completed,
returned questionnaires decreases. The average proportion of com-
pleted, returned questionnaires to the number of questionnaires
handed out was highest at Harrisville State Park (low-intensity usage).
Pinckney State Recreation Area (intermediate-intensity usage) had the
next highest proportion and Dodge Brothers #4 State Park (high—
intensity usage) had the least number of questionnaires completed
when compared to the total number of questionnaires handed out.

Of the total number of questionnaires returned, approxi-
mately one-quarter were incomplete. Therefore, they were not
included in the analysis. Depending upon the degree of rigor used in
judging completeness, from 10 per cent to 25 per cent of the total
questionnaires returned in similar surveys will probably be lost from
analysis.

The percentage of questionnaires returned to number of

questionnaires handed out varied from 48 per cent to 32.5 per cent

106

(excluding Test 8-—the mail-back test-—in which 15.5 per cent were
returned). As the testing schedule progressed, the average number
of returned questionnaires decreased slightly. This may be due to the
length of the testing schedule. Since tests were conducted every week-
end for five consecutive weeks, there is a good possibility that some
park visitors received two or more questionnaires during the testing
period. If an individual completed the first questionnaire he received,
it is likely that the motivation to complete additional similar ques-
tionnaires was less for each succeeding questionnaire.

The proportion of questionnaires completed to the number
of questionnaires handed out varied between 31. 5 per cent and 24. 5
per cent (excluding Test 8——the mail—back test--in which 15.5 per
cent were returned and completed).

The results of the significance testing of variables are ex-
pressed in Table 7. All calculations were made at the 95 per cent con—
fidence level following the procedure outlined earlier in this chapter.
Of the total number of variables tested, only three alternatives exhib-
ited a variance which was significant at the 95 per cent level of con-
fidence. However, several other variables approached this signifi-
cance level. The general conclusion one might draw from this is
that park users appear to be of two general types when considering

questionnaire completion: one type of park user will complete a

107

.coBmEommU ho.“ um momma oomo

.omaommop mo mmouopmafioo
.Hom cogfimxo who? mmfiwccofimosq @2338 mmmfi .mummp “nosvomnsm 5 .ouofldfioo omnmgmmoo
who? 3 “won. 5 @2238 mmpwmccofimoao So “any «03 2: Song :Smop m 5 mooconmmfimo—

.m 2an o“ ammo.“ ofisoonom ammo. Umfimumo hobo

 

 

 

 

 

. mum; 30:
m5 3mm pm: mum; 0
Z
$-73
oz .m> .Q 3
wow. am pm: MHMHM 30:
Q MYTH 69:on
wnmufi
. mum; Upmo
mww m3: mfimm NINIH omgacb
TN;
02 .m> .< H
TH;
. m-3-3 nnmo
m5. 002 mHmm NIH; 6039.3
3-3-3
3.63 3.8 m .8 8.33800 So 8933 8383383
«m o mmfiwccofimosg mmfiwccoflmosg . o3m€w> pooh.
:oEmHDonU mfiwcnofimosmv m
mocmowficmwm go $5852 mo .HQQESZ

 

 

 

 

 

 

 

 

$335.5 .afiqmfitmaxm: mo 3:83. we. .8 $3QO -- 3. mamafi

108

 

 

 

 

 

 

mu3q- m 33 3:35
wmm. mam S: NATN £953
HLVIN Undo menH
oz .m> m
fimcfl
pom. Q33 gm N33“ __.3w3:wmm__
vmm. mm3 mam Nu3nm .3630
073 www. wm3 mam 3-3-m cofimmsv w
3&3». N: mam mn3um 3538.323
3033
wow. m33 gm N333 073
oz .m> m
2033
0mm. 3&3 gm 3-3-3» 3530a
3963 030mm 333800 30 nmccmm
m mm .30 mmp3m§o33mms~u mmfimccofimmsg 330330333383 m3£m3pm> 59H.
3 3303335030 . . . whmccofimmsmv
wucwowflcwfi . .30 umnanz .30 3933333373

 

 

 

 

 

 

 

 

UmfiGECOU .. n . N. WmewH

109

 

 

 

 

 

m m m
. mumum c.3393
3&3 3mm oom3 3umum xown 332
w Ewe m
mm? .m> um
. 3 u 3.. m 93:36.3 3.
3% 3% ooo3 3.3.@ :3 3:333
3. 593.
um 38
m-m-m 3 n 3.
. u .. cam :3 USE:
m3m mom 3m33 MIMIM 5953
v.30 pmom
073 .m> m
. 33 3 w 5983
now m 5m N :pmngm:
3393 $8 m 30 832980 30 @3333 83833383
3m mmfiwcaofimmsg mmfimccoflmmsg . @3333? $93.
33033335030 ohmgoflmmnG
mocmnvwficmfi 30 393333573 30 3238373

 

 

 

 

 

 

 

 

3.25380 -- .3. @3333.

110

 

3303333

 

 

 

 

mmm. 33 com 733-3-m oZ
mew .m> m
. Ummo3ocm
23 2% com 33.3.3” :0qu
Wm E
o3. $03 32.. N33 Emu
3-3-3 :35:
mmW .m> 3
Hum
$3. $33 82 313-3 3:8
T: .3333:
333 $3 .3 30 832388 :5 822; 8383383
3w mmfimgofimmsg mmfimccofimmng m3nm33w> KPH
no 3333630 mfimccofimmsG
mocmoafima .30 3938573 .30 3933333373

 

 

 

 

 

 

 

 

UmSGEGOU u: .3. HJm<H

111

 

 

 

 

 

 

 

 

wwm. mm3 mam mumum maoflmmnv
o mam. N33 Sm wimum m3mh3wcw
. mm Mm mm m... .
amum 3
333. 33 gm mumum 3 33.33
Sm. 33 SN TTm mmqfio
wmm. $3 mpm mu3um 333%
. cm xooam
own mm3 mrm mu3um .3. pmmmm
OZ .30 3V
mum $3 mpm mn3um agonm 30300
Em. mm3 3h 33$ 9:55
mwm. mm3 mpm mu3nm 923m
3263 .358 m 30 863.3800 30 82333 85833282
3 mmfimficofimmng mmpfimgoflmmsg . m3nw33m> «mmH
383833830 9339338335330
mocmomamfi 30 3235373 30 pmngz

 

 

 

 

 

 

 

 

3305:3380 .1. .3. Mdmdfi.

112

questionnaire without being unduly influenced by the appearance and
content of the questionnaire itself; the other type of park user tends
to ignore the questionnaire in spite of different changes made in ques-
tionnaire appearance and content.

Based upon the results of this analysis, the questionnaire
that would obtain the largest amount of response would have the fol-
lowing characteristics (ranked in descending order of importance
with regards to questionnaire completion):

1. The questionnaire would be handed out at the entrance
to the recreation area and would be collected by means of a deposit
box at the recreation area exit.

2. A pencil would be supplied with the questionnaire in
order to facilitate completion.

3. The questionnaires would be printed on a "heavy" card
stock.

4. The color of the questionnaire should be, in order of
preference: blue, white, brown, tan, yellow, and orange.

5. The questions concerned with activity analysis should

be constructed according to the following preference ranking:

2-—3-—5, 2--3--2, 2——1--1, 2—-3--4, and then 2--3--3. From these

 

1See Figure 11 for illustrations of the activity analysis
questions.

113

results it appears that a closed, structured approach to collecting
activity information would tend to obtain the largest amount of com-
pleted response.

The small difference in return between the regular length
questionnaire and the post card length would indicate that as long as
the questionnaire appears to be short in length, response will not be
affected.

When a "voluntary deposit" retrieval system is used in a
park, the provision for pre—paid mailing of the questionnaire after
leaving the park appeared to decrease the number of questionnaires
returned. This could be due to the fact that respondents saw that
they had an alternative way of returning the questionnaire. (It does
not necessarily have to be completed while they were in the park. )
The questionnaire was then carried from the park and, in spite of
good intentions to complete and return it, the card was then forgotten
or mislaid.

The inclusion of a slotted hole in the questionnaire did not
appear to increase the response rate enough to warrant its additional
expense.

The results of the analysis of variance and chi square in-
vestigation of Experimental Test 2 (determination of "best" activity

analysis question) indicated that there were no significant differences

114

in the responses to the five different questionnaires. This compara-
tive analysis was conducted to determine whether particular ques-
tionnaires ”attract" responses from different segments of the park-
using population. It appeared that despite the content and appearance
of different questionnaires (if the compared questionnaires were
reasonably similar), the populations completing the questionnaires
possessed similar characteristics.
Sub-hypothesis one is stated as follows:
Through a process of trial and selective elimination, a
questionnaire and dispersal-retrieval system can be designed
which will obtain a greater number of usable completed re-

sponses than other questionnaires tested under similar condi-
tions.

The results of this study phase indicate that sub-hypothesis
one cannot be rejected. Testing of ten variables which might have
affected response proved that by changing questionnaire design, con-
tent and dispersal-retrieval, response rates can be altered. On the
basis of the experimental testing of each variable, it was a simple

task to choose the particular alternative that received the greatest

number of completed responses.

CHAPTER VI

RECOMMENDATIONS

The recommendations that follow apply generally to

recreation survey research. Specific recommendations directed

to the Michigan State Park system are included under a separate

sub-heading later in this chapter.

General Recommendations

 

1. Continuous data collection systems should be imple—
mented in all park and recreation areas. As the number of potential
recreation sites declines, and the use-intensity of present sites
increases, more pressure will be brought upon park and recreation
agencies to provide leisure opportunities. This increased pressure
is already becoming apparent in the general interest displayed by the
public towards park and recreation agencies, e. g. , public interest
in the use of a recent Michigan recreation bond. Under these con-
ditions, the park planners and administrators must have accurate
data on which to base the decision-making process. Usually, the

most valuable data is that which reveals "trends" in user

115

116

characteristics and characteristics of facility usage. This type of
information probably provides the best present indication of probable
future demand.

2. When recreation area "day-use" information is desired
(on a continuous basis), a questionnaire data collection system of the
type described in this study should be used. At the 95 per cent level
of confidence, information collected by voluntary deposit of question-
naires showed little significant difference from collected parameter
information. The one area in which a small significant difference was
noted was in the answers to questions concerned with travel time and
distance. In spite of the inconsistencies noted in the answers to the
travel questions, one trend was detected. Information collected by

voluntary deposit of questionnaires showed slightly more ”travel
time" and "miles traveled" than did parameter information. In
future questionnaire surveys when travel information is desired,
the respondent should be asked for both "mileage traveled" and
”travel time. " Based on the results of this study, if the response
to one of the questions proves inaccurate, responses to the other
travel question will likely yield the desired information.

A questionnaire data collection system similar to the type

described in this study can usually be implemented for a fraction of

the cost of a comparable interview data collection system.

117

3. When "day-user" information is to be collected by means
of a questionnaire, the questionnaire should be handed out at the area
entrance, and collected by voluntary deposit at the area exit. This
type of questionnaire dispersal-retrieval was more than twice as
effective as a similar "mail-back” system tested.

4. All aspects of the survey system should be coordinated
so as to provide a central theme or "motivational continuum" for the
research project. If the same colors, language, survey symbols
(e. g. , silhouette of a picnic table) can be used on all components of
the survey system, the opportunity to associate components is in-
creased. It is believed that the coordination of survey components
was a major reason for the relatively high proportion of questionnaire
completion in this investigation.

5. When designing a questionnaire and dispersal-retrieval
system, every effort must be made to avoid possible non-response.
Based upon the results of this analysis, the questionnaire that would
obtain the largest amount of response would have the following
characteristics (ranked in descending order of importance with
regards to questionnaire completion):

3. A pencil would be supplied with the question-
naire in order to facilitate completion.
b. The questionnaires would be printed on a

"heavy" card stock.

118

c. The color of the questionnaire would be, in
order of preference: blue, white, brown, tan, yellow,
and then orange.

(1. Questions dealing with analysis of activities
would be highly structured. Answer categories would
be "closed, ” perhaps in the form of "simple" graphs.

e. The questionnaire should "appear" to be of
short length. The amount of material contained in a
questionnaire does not seem to affect response as much
as the appearance of being brief.

6. Better guidelines on the precision, reliability and
validity of recreation research should be established. It is sug—
gested that future recreation research be conducted at the 95 per cent
level of statistical confidence. When calculating sample size, the
estimation error should be no greater than .05 at the 95 per cent
confidence level. Unless a concentrated effort is made to raise the
standards of research within the field, even less attention will be paid
to future research than has been the case in the past.

7. Whenever a data collection system utilizing a question-
naire is being conducted, a further test (by means of the interview
technique) should be made to determine the accuracy of collected

information. In most circumstances, the amount of effort involved

119

in an additional interview accuracy-test is worth the trouble and cost.
Too often, collected data is accepted as being 100 per cent accurate
when this is in reality not the case. All data has limitations and if
these limitations can be carefully and concisely outlined, the col-
lected data can be used with more confidence.

One interview-test per season should be sufficient to ascer-
tain the accuracy of information collected by questionnaire. This is
assuming that there are no drastic changes in the questionnaire con-
tent or design, the questionnaire dispersal—retrieval, the park itself,
or in the population using the park.

8. Further investigation should be made of the I. B. M.
optical scan process. This method uses the reflection of light from
a heavy lead pencil mark to produce a data processing (computer)
card. Theoretically, any base material that does not also reflect

' If a questionnaire

light could be used as a questionnaire "base. '
could be designed that could be read electronically, with no inter-
mediate coding stage, survey research could be performed at a
great financial saving.

9. Whenever possible additional staff members should be
used to conduct survey research. It is suspected that one of the main

reasons for the high proportion of returned questionnaires in the low-

use parks was that the park staff member could devote more time

120

to "selling" each potential respondent on the importance of the
research. Additional staff members, with sole responsibility to

the survey research project, could take the time for an enthusiastic
motivating message without worrying about the other responsibilities
of a contact station ranger. Also, special staff members could be
trained in research methods, thus helping to ensure the validity and
reliability of collected information.

10. Survey researchers should give more attention to the
determination of the method of data collection that best meets their
needs. As more expertise in research methodology develops, and
as more financial support is made available for research, this prob-
lem will likely be eliminated.

11. Further research needs to be undertaken to check the
validity of the collected information.

Recommendations to the
Michigan State Park System

 

 

1. There appears to be a growing trend to use state park
and recreation areas throughout the entire year, rather than merely
for the short summer season. In order to ascertain use and user
characteristics of this segment of the park—using population, con-
tinuous data collection systems should extend through the entire

year.

121

2. The entrances and contact stations in the majority of
state parks do not lend themselves to a user survey. There are
usually no distinct separate entrance lanes for vehicles with season
entry passes or for those "vehicles" purchasing daily entry permits.
Traffic congestion makes traffic survey research difficult. All
vehicles entering the park should come to a full stop. On the other
hand, if a "vehicle" possessing an entrance permit can drive directly
into the park without having to stop, any type of sampling or survey
procedure is difficult. At present, the task of the contact station
ranger is difficult even when a survey is not being conducted. When
a Visitor is at the contact station purchasing a permit, the ranger has
to virtually ignore the other vehicles driving directly into the park.

3. More attention should be given to the construction of
adequate contact stations at park area entrances. If there were two
entrance traffic lanes and the contact station was located between
them, more control could be exercised over incoming traffic.

Better communications between the park administrative
office and the contact station should be established. In most parks,
if any problems arise at the area entrance the ranger must leave the
contact station to render assistance. Passing motorists must be
depended upon to notify other rangers. If the door to the contact

station was located either in the front or side, the ranger could

122

leave the station without losing sight of the traffic or the front
counter (and his cash box) in the station.

4. More effective means of obtaining counts of all vehicles
entering the park should be installed. A number of state parks do
not have an accurate count of the number of vehicles admitted.
Determination of sample size depends upon an accurate count of

vehicles entering recreation areas.

SE LE C TED BIBLIOGRAPHY

SE LE C TED BIBLIOGRAPHY 1

Books and Periodicals

 

Bachrack, Stanley D. , and Scoble, Harry M. ”Mail Questionnaire
Efficiency: Controlled Reduction of Nonresponse. " Public
Opinion Quarterly, XXXI (January, 1967), 265-271. (Q)

 

Backstrom, Charles H. , and Hursh, Gerard B. Survey Research.
Evanston: Northwestern University Press, 1963.

 

Bauer, Raymond A. "Social Indicators and Sample Surveys. " Public
Opinion Quarterly, XXX (February, 1966), 343-345. (C)

 

 

Cannell, Charles F. , and Fowler, Floyd J. "Comparison of a
Self-Enumerative Procedure and a Personal Interview: A
Validity Study. " Public Opinion Quarterly, XXVII (1963),
250-264. (C)

 

Clawson, Marion. Land and Water for Recreation: Opportunities,
Problems, and Policies. Chicago: Rand, McNally and
Company, 1963.

 

 

, and Knetsch, Jack L. Economics of Outdoor Recreation.
Baltimore: The Johns Hopkins Press, 1966.

 

Cochran, William G. Sampling Techniques. New York: John Wiley
and Sons, 1964.

 

 

1The letter in parentheses immediately following research

reports refers to the research method used to obtain data, i. e.
(O) = Observation, (1) = Interview, (Q) = Questionnaire, and (C) =
Combination of research methods.

123

124

Colfax, David J. , and Allen, Irving L. "Pre-coded vs. Open—ended
Items and Children' 5 Reports of Father' 3 Occupation. ”
Sociology of Education, XL (January, 1967), 96-98. (C)

 

Deming, W. E. "On Errors in Surveys. "

Review, IX (1944), 359-369.

American Sociological

 

Doxiadis, Constantinos A. Emergence and Growth of an Urban
Region. Detroit: The Detroit Edison Company, 1967.

 

Festinger, Leon, and Katz, Daniel. Research Methods in the Be—
havioral Sciences. New York: Holt, Rinehart and Winston,
1966.

 

 

Good, Carter V. , and Scates, Douglas E. Methods of Research.
New York: Appleton-Century-Crofts, Inc., 1954.

 

Gorgen, Kenneth J. , and Back, Kurt W. "Communication in the
Interview and the Disengaged Respondent. " Public Ojlinion
Quarterly, XXX (Fall, 1966), 385-387. (I)

 

Kerlinger, Frederick N. Foundations of Behavioral Research. New
York: Holt, Rinehart, and Winston, 1967.

 

Knudson, Dean D. ”Response Differences to Questions on Sexual
Standards: An Interview-Questionnaire Comparison. "
Public Opinion Quarterly, XXXI (1967), 290-297. (C)

 

Mendenhall, William. Introduction to Probability and Statistics.
Belmont: Wadsworth Publishing Co. , 1967.

 

Moser, C. A. Survey Methods in Social Investigation. London:
Heinemann Educational Books Ltd. , 1958.

 

Parten, Mildred. Surveys, Polls, and Samples: Practical Pro-
cedures. New York: Harper and Brothers, 1950.

 

Reissman, Leonard. Class in American Socieiy. New York: The
Free Press, 1959.

 

Riley, Matilda White. Sociological Research: A Case Approach.
New York: Harcourt, Brace, and World, Inc., 1963.

 

125

Selltiz, Claire; Jahoda, Marie; Deutsch, Morton; and Cook, Stuart W.
Research Methods in Social Relations. New York: Holt,
Rinehart and Winston, 1967.

Sessions, Frank Q.; Epley, Robert J ; and Moe, Edward O. "The
Development, Reliability, and Validity of an All-Purpose
Optical Scanner Questionnaire Form. " Public Opinion
Quarterly, XXX (1966), 423-428. (Q)

Public Documents and Reports

Chubb, Michael. Outdoor Recreation Planning in Michigan. Report
for Michigan Department of Conservation. Lansing: Michi-
gan Department of Conservation, 1967.

Clawson, Marion, and Knetsch, Jack L. "Recreation Research:
Some Basic Analytical Concepts and Suggested Framework
for Research Problems. ” Proceeding of the National Con—
ference on Outdoor Recreation Research. Ann Arbor,

Michigan, 1963, 9-42.

Detroit Metropolitan Area Regional Planning Commission. Home
Survey of Regional Recreation Activities. Detroit: Detroit
Metropolitan Area Regional Planning Commission, 1959. (Q)

 

Park Users Survey: A Survey of Persons Using Eight
Selected Regional Parks in the Detroit Region. Detroit:
Detroit Metropolitan Area Regional Planning Commission,

1959. (Q)

Hirsch, M. Lower Fort Garry National Historic Park Visitor Use
Study. Canada: Department of Indian Affairs and Northern

Development, April, 1967. (O)

James, George A. , and Henley, Robert K. Sampling Procedure for *—
Estimating Mass and Dispersed Types of Recreation Use
on Large Areas. Asheville: U. S. Department of Agricul—
ture, Forest Service, March, 1968.

 

James, George A. , and Taylor, Gordon D. ”Recreation Use Esti-
mation in the United States and Canada. " XIV IUFRO-

KONGRESS. Section 26. (Munchen, 1967), 15-29.

 

126

King, David A. Characteristics of Family Campers Using the Huron-
Manistee Natural Forests. St. Paul: U. S. Department of
Agriculture, 1965. (I)

 

 

Mutch, W. E. S. Public Recreation in National Forests: A Factual
Survey. Forestry Commission Booklet No. 21. London:
Her Majesty' s Stationery Office, 1968. (I)

 

New Jersey Department of Conservation and Economic Development.
New Jersey State Park Visitor Use Survey 1966-1967. State
of New Jersey: Department of Conservation and Economic
Development, 1968. (I)

 

Nixon, H. N. Banff National Park: Aspects of Visitor Use. Canada:
National Parks Service--Planning, Department of Indian
Affairs and Northern Development, March, 1967. (I)

 

Prince Albert National Park Visitor Use Survey. Canada:
National Parks Service--Planning, Department of Indian
Affairs and Northern Development, January, 1968. (I)

 

 

RidingMountain National Park Visitor Pattern Survey.
Canada: National Parks Service--Planning, Department of
Indian Affairs and Northern Development, June, 1968. (C)

 

 

North Central Texas Council of Governments. A Parks Appraisal.
Texas: Texas Technological College, School of Agriculture,
1968.

 

Ontario Department of Lands and Forests. Ontario Park User Survey-—
Methodology and Results. Toronto: Department of Lands
and Forests, 1967. (I)

 

 

Outdoor Recreation Resources Review Commission. National Recrea-
tion Survey. Study Report 19. Washington, D. C.: U. S.
Government Printing Office, 1962. (I)

 

 

Outdoor Recreation Literature: A Survey. Study Report
27. Washington, D. C.: U. S. Government Printing Office,
1962.

 

 

Participation in Outdoor Recreation: Factors Affecting
Demand Among American Adults. Study Report 20. Wash-
ington, D. C.: U. S. Government Printing Office, 1962. (I)

 

 

 

127

The Quality of Outdoor Recreation: As Guidance by User-
Satisfaction. Study Report 5. Washington, D. C.: U. S.
Government Printing Office, 1962. (C)

 

 

Taylor, Gordon D. Techniques in the Evaluation of Recreation Use.
Ottawa: Department of Indian Affairs and Northern Develop-
ment, 1967.

 

 

Thorsell, J. W. A Trail Use Survey--Banff and Yoho National Parks. L/
Canada: National Parks Service--Planning, National and '
Historic Parks Branch, Department of Indian Affairs and
Northern Development, February, 1968.

Waterton Lakes National Park Visitor Use Survey.
Canada: National Parks Service--Planning, Department
of Indian Affairs and Northern Development, March, 1968.

 

U. S. Department of Commerce. Tourism and Recreation: A State
of the Art Study. A report prepared by Arthur D. Little,
Inc. , Washington, D. C. : Government Printing Office, 1967.

 

 

Other Sources

 

Crane, Dale A. "Progress Report, Presented to Conference on
Methods of Demand Analysis for Fishing, Hunting, and
Other Outdoor Recreation Activities. " Washington, D. C. ,
February 27-28, 1967.

Jackson, Robert Morton. ”Differential Value of the Mailed Question-
naire and the Interview in a Follow-up Study of High School
Graduates. " Unpublished Ph. D. dissertation, University
of Wisconsin, 1959.

Shafer, Elwood Lewis, Jr. "Effects of Sampling, Location, Period,
and Method on Camper Survey Results. " Unpublished Ph. D.
dissertation, State University College of Forestry at
Syracuse University, 1966.

APPENDICES

APPENDIX A

”SEASONAL" QUESTIONNAIRE

128

 

 

 

 

 

"Hakka “mmgbz HQOO vzgd

 

 

 

6qu EB undo—Auntflon 3E3 «on on

olnl-I-lol-l

"Ha—gm..—
.I all. 1 .III' ...III .| "NA;

«ma—0333:, Emu. Zn NEONA— firh ho a0< HE Hfi< B<§ V

 

35 SE;
.550 D .3.
3:2,: 30 965 D .m
one? 520.... 23 D .3

33:00 0.58 no 28 D .38
33.0 2960 one D .0
To... .ezo .53 .485. 8.650 55 8553 25 D A.
95.3» 3362:de D .0 £339 533 .3358 one D .a

88 “.85.
... “So—=3; ma. 5 $93 mzr flange ham 0253.393 mi. ..30 53:3 m

 

2. D 8.3 D 338an gm 8:. E 458 2. 968 no» mm... N

 

...... ..
...SD. EU «Eggumémzaagopowan —

 

....33 x52 my; m<mz x8 NE. 23 Bio .31»
m><m4 g corpmwba rumbw .50 .33... mm<m3m
53.555 8.8 4.965 B 8:29—80 HERE 3 2&5
9.305960 05 p5 coautomnoo ”gig of 5 6326.000 v.33 2 it 35.
53.338... 38.5 B 5.;
Danna so» 33: coaaEL8E :a 62:00 36 .3 lie 3% In :33 33 58>
wmhazE no H.909 1 >420 3:2. ..3 ; ....
$280.28 263335 :a 50 333.3 0305
330.38 36% 353333
05 83.5.3 05. E8»? fan .335 293:: 223 3 van: on :33 who: 9:.» 30.3 con-35.385
23.3. 33.3.5 35m .336 on: on? 0:300... 05 33930.3 3 3.3023 0.35 no one 0.3 :0.»

 

..tl. l4 4 «J. 1.1. 4 1
. .. .
. . . .. . 1

Wmmx..s...r..w.1 . . . : . .. ......
i g... 5.: a * ...... . . . . ...........U..m,t¢f..

I. .... . p . £3523

. 2.4.... .

 

 

 

8.. Q E... 5 so 2 use. afiwn E...

”88 93305 6535 in
.5285: 38m 558.:

95:5 88.58: 3.282 .2” 58m
:5 9.35.4 v5 £83833 832833

63 s 33.2: 0303.3 Jena 05 89c .32: 6.30 «35 cor—.30 5353.303 023 so» 33

mic: as: mug nz< 59353.3 z< m><= Ed: 55> mo“. 92$:

 

 

 

 

 

3:23 .3...

932. .0

9:6: ..3

p.32. .a
a 083.3. 35353333

 

.EEU< NE (.30 EDIE 28 9.2QO memo EDD? E3033
ho mag MHZ. zu— flg 0.0434 .28 Han. 88: NE. Emma 990330
~50? EU< NDOh "Hm—.3. 2— NE .BSflm mafia—trauma mflo<mm N28 23 VN

 

 

.5 3:5
.560
. 8335. D .u
Eggs—U u 353$ 4:33 D ... was... .
$53.8 Di 3...: .35.. D .o D 3
93335530 D .o
c.3830 0.381: .n
.8 4850385338.» __ unconD

. 9503??— .E .3393 a .33

P33333650 s cm .— 3.33.3255:-
aa: 5...:- SEE .. 9.383 D .u
~85... 8 wagséD .. 3:1. D ... 9.328% D .0

5: 388.530.. 5.35% D .3 358
can: :5 D ... .35 028. 3 48:3 D .n

3.3% 33a .5 a]. D .3 3333 .30

gaaaé 9.31.8330... apnnatoagon—J

:33. as: .83 ea :- 388. 3.53m:
mg on gone .50» Ba 33: 83.5.3 E3 may. no 3553 MN

 

142.625. .. /,

 

.91- .l. .. . . . .... . . .. .
“Mk. WW3»: ir ... Sire...“ . ...-m... Lain}. ..l ...—rust.

 

.0

 

.n

 

«g..— nflh. 8 ONCE nun 0.3. EA DOV 383 EEOSH
GO EBMOAn—O ZE<§OH¢ flab—Ga ho RE RE NN

 

 

 

APPENDIX A

"SEASONAL" QUESTIONNAIRE

Unfolded--Showing Front and Back Pages
' (Actual Size 8% Inches X 11 Inches)

129

 

ES: .3 D ...

93E .3 932. Nd D .u 993 env D .o
952. «72 D a .59. Ta D .p
932. Sum D .o 32 3 932. n D .a

38 :85. $552. 523 m5. 3 92mm... :0» Be 950: g 30: —N

 

...Lxm x¢<a 9:. «(m2 Qma_>omm xom 5:. 2. 9:3 m—rh m><m4
o... .395“. .328 Qz< ..Lm; mno> >032m .xmf. NIP w><w4 30>
mmOuwm kmnfi 4:2: mzo_._.wm:o oz_z_<2wm MIA. .50 4...... .82 on

 

 

s mzo gamut—a < 25:. 5:93 3:24 BE. $003 so.» an E on

88.8586» Dd
3.6 Ba 88.3» D.» 88.8588 D .o
m8.§».o8.2» D a «8.388....» D .n
m8.3».o8.2» D6 88.3 5.5. D .n
A08 185. «in: 35 g 45.2.
mm... 8308a 5mm 03342 mi. ..3 .353 «—

ouoE no S o— m.— v~ nu

D D U U D
NH «a an a m b m n v m N a
DDDD DDDDDDDD

 

 

.5... 86.8”“ D .u

.Ed 85785 D .u .5... 8&é8a D 6

.Ed 85.86 D c 82-8.2 D .9

.53 36.6ch D .0 .58 85783 D J
38 x83. ..me a5. 552m :0» So was 955 m—

 

 

 

D
D

D

32952 «—

 

mefim 2

 

72:33:52
Ea E30 9—

 

DUDE]
DUDE]

U
D
D Eu 3
a

 

D
D D U
D D D
D D D
a a clam

o
a
D

Z

unau on as munav
uo>o onlfin own: 076 nn~

.Smhmhm xm<m 20<m mom m§m2< H20 v—Omzo «mm:
a 38mm ENPE 259 mbO—m<> 8:. Gum; DO? m><x wkcd >22 30:

32:“;
£255 one :85. 9.555 53» .3 9.3:... 8:. .3
33.3200 22.5028 3 m5: ..3 xmmzpz 455,—. . 8:. .3
m5. 8:28.: 3mm 33mm Ewamzc. m2... .3 353 N—
2: 833. 833.600 away—Q
5:. .3
a 02:5
< mo... on ...SEE :3» .3 35:... E: 88 SE? —— nmmmamz...
03:3. D .
Sam Bag .......... ... mo< mm 8.
22: D .
s ...».Bfim E3» .3 9.5::
mm... .3 mo< mséuxoxmfi 92¢. xmm mm... a 955 e—
00:: uuuuuuuuuuuuuuuu
p 83:

30.» 29: 3.3 9:. a .mesom sumac 302 m5. 3 .835 >24: 30:

 

 

n3 :3: n.— :6:

Tax? NIH. OZS< NEH. 835/2089. OEQDAOZ— EOE
w><QOF Vilma ETC. OF GEAM><¢E 929nm 30% 8Q EB :02: 30:

 

nnnnnnnnnnnnnnnnn N. MDOU .m—N £20.» 9 95.—3

 

33m 3530 .05 .3 .58.

 

62.82 3: 8223 «8.: 83.3 s 3.3: 3.meme E3» a 38.3

 

 

 

:3 3 .33 D A :3 ~74. D .u
8% 8-; U .u 25 a-» D 6
£96 8-: D .u 9.26 T" D .a

:3...
1% 2-2 D .o {a 3.5 as 3 25 D ..

“mam” a afim i=8 Dumb DO.” H><2 m><n >22 30: ‘—

 

2. 3:3.
.343 D .m
9.333., .393...
3:32
a u
and 3.8 32:25 0.35 .3 9.5 3 «.3 D 0
005353 a 3 tan D .o 35.3:
2.5 335 .338 a no En D a
53:33 5 3 tan D .u 3.5 .3 33:0 and 93 D J

A23 3039 «ADOmU KEG? mow W29 m Emir ho BAU— aka—3

 

 

- - Continued

APPENDIX A.

APPENDIX B

WEEKLY REPORT FORM USED IN

”SEASONAL” SAMPLE PARKS

130

down; Hwfio mmbv nonHqumem mans

“503$..." 98 95» «>2: 95qu ME;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

2H mnm<o sac mom¢u Illub¢||uml.uq 2m
.2H mom¢u goo mamau .Ilmxcmulunlnlq z< z=m
2H mam¢u aao mamcu lulww>¢ a a 2m
, zu.mam¢u ‘eoo mnmgu u>4 m a 24‘ wa.
.zH mamao .aao mam¢o w>4 m a 2m
.zu.mnm¢u apo.mnm¢u w>4 m g _a<; .mm
L E W25 So 858 92 m A In.
. ; om4o Boo mom¢u u>4 m a :< «zzp
2H mammo, .933 nom<u o>¢ m 4 Arm
2H woman aoo momcu u>4 m a z< am:
...-.45 858 So man I 1 92 m a z“.
zurmam4u sac mango u>4 m a xz<
t 2H mom¢u aao mam4u u>4 a q :a
2H wom<u sac mamau o>< . m A zqa.
. auounccaz .¢
...I .coo: nouns nuufidu 52.6 .n .. you: .q amouuflwo .m
.3 mmuuo 05 can soon 0.3qu 3098 .. 3330: .m >258 m>¢uommuo>4 .m
co>wooou nouco no nonauc uuooom yam“; .u uozonm ago“; .~ aauumm .~ munmam .~
2.02 2" ocoz .H 53.8 ..n .733 ...”
.600: uouuo can :00: 3800 v.33 counumuwmwooum nm>ou om ummumoc >43
“50 03.3: nun—do no none—s: vacuum 050.8

 

 

”9.3 on ousuaummaoa

 

m~_m m4¢z<m

 

mzo_»_azou mmzh<mz

 

 

Hmommz xm<¢ >4xmmx 4<_um¢m

APPENDIX C

"SHORT FORM" QUESTIONNAIRE USED

AS THE INTERVIEW SCHEDULE

131

 

010 YOU ow m ms PARK LAST mom?

 

[] Yes [:]No
AREYOUD some TO ow IN THIS PARK TONIGHT?
Yes
UHAT ARE THE AGES OF THE PEOPLE IN THIS VEHICLE?
MALI: ____l -.—.-. ______I ‘___.D _____0 _____'
:L‘IALE: -_‘_, ‘_ . , , ,

 

HOW MANY HOURS DID YOU SPEND IN THIS PARK TODAY? (Check

a. [J 2 hours or less
b. C] 2-4 hours
c. [3 4-6 hours
d. C] 6-8 hours

 

e. [3 8-10 hours
t. [J 10-12 hours

3. [J 12 hours or more

One)

 

 

WHERE IS roux PRESENT HON—E? (Exact street address not requireay

 

Town or City I County State

 

WHAT IS YOUR ZIP CODE?

 

HOW MUCH TIHE DID YOU SPEND TRAVELING TO THIS PARK TODAY!

(NOT INCLUDING "STOPOVER" TIME ALONG THE HAY.)

Minutes

HOW MANY DAYS HAVE YOU USED THE PARK IN 1968?

a. [3 this is the first park e.

visit
b. C] 1-4 days
c. [3 5-8 days
d. E] 9-12 days

h.

C] 13-16 days
[3 17~20 days
[3 21-24 days
E] over 24 days

 

 

HOW MANY MILES. BY THE MOST DIRECT ROUTE, IS THIS PARK FROM
HOME?

Miles

 

 

 

HHAT IS THE SEX AND APPROXIMATE AGE OF THE
"HEAD OF YOUR FAMILY"?

[3 male
9 .
AGE. __ ________ years SEX. [3 female
WHAT DOES THE "HEAD OF YOUR FAMILY" DO FOR A
LIVING?

 

WHICH OF THE ANSWERS BELOW BEST INDICATES THE
TOTAL NUMBER OF YEARS OF EDUCATION COMPLETED
BY THE "HEAD OF YOUR FAMILY"? (Check one answer)

DDDUDDDDDUDU
DDUDD

KHICH OF THE FOLLOWING BEST DESCRIBES THE
TOTAL INCOME OF YOUR FAMILY LAST YEAR (Check one)

WHICH OF THE ACTIVITIES LISTED BELOW DID YOUR GROUP DO

HHILE HERE? (Check all the boxes that apply).

e.
fl
g.

E] sightseeing
from car only

[3 walking to
scenic points

[3 picnicking

[3 looking at
plants, animals
or birds for a
hobby

[] swimming

E] wading
[J sunbathing

OTHER

h.

1.

j.
k.

E] water q. [3 games 8
skiing team sports

C] skin or r. [j trail hikingl
scuba diving 3. [] horseback

E] notorboating riding

[] sailing t. [J listening to

1. [] canoenng

ll.
n.
0-

p.

[j rouboating
[j boat fishing

[3 bank fishing museums or
nature
[3 fishing
(wading) centers

ranger talks

u. E] taking
guided tours

v. [j visiting

u. [j relaxing
x. [3 photography

 

(write in)

 

a. 0 under $3,000 a. B 510,000 - $14,999
IL [J $3,000 - $5,999 f.[] $15,000 - $24,999
c. D $6.000 - $7,999 g. 0 $25,000 - and over
d.C] $8,000 - $9,999

 

WHAT TIME DID YOU ENTER THIS PARK?

a.[] 8:00 -10:00 a.m. (Check one)

t>. C] 10:00-noon e.[] 6:00-6:00 p.m.

c. E] noon-2 :00 p. m. f.EJ 6:00-8:00 p.m.

d. [:1 2:00-4:00 p.m. 34:] 3:00-10:00 p.m.

IN THE SPACES PROVIDED BELOW, WRITE IN THE FOUR ACTIVITIES
YOUR GROUP SPENT THE MOST TIME DOING.
NUMBER OF HOURS YOUR GROUP SPENT DOING EACH 0? THE

ACTIVITIES.

Activities

ALSO WRITE IN THE

Time Spent

 

 

 

THANKS FOR (OUR HELP:

HAVE AN ENJOYABLE & SAFE TRIP HOME

 

LICENSE NUMBER

 

 

PARK CODE NUMBER: '1 DATE: 1

IDENTIFICATION NUMBER

 

 

APPENDIX D

OPTICAL SCAN FORMS USED IN CODING

132 -

 

 

 

 

 

 

 

 

 

I I? I N 1 I P .' 1’ A T I U N I N I N F ‘
‘ I I 1 3 C S C 1 I 9 3 g g a 3 a g g 1 g g
‘ o 1 1 1 a s s 1 O I ‘ a 1 1 4L 0 L s 1 L 1
‘ g 4 J 1 1 I 1 1 1 1 l x
I L 1 1 1 1 4 1 1 I 1 ’ 0 1 1 L A j o 1 1 1
1 4 1 1 a ’ 0 j I 1 O I 0 1 3 9
: 'vm
M
(I'D WTWVIIT IV 7 :1uw—r III-I iii-T Tlun‘rr
lerwr VII-r 1v—a‘w‘rr ' "I—"II" T‘I VIUV'
# l A . l . L I . I
t I ti 1: s - n I 1. rt "
1 -’ 0‘. OI, 0’. I.’ 'u '0’ ‘0. 'D. “0 3 ‘D ~10 'E‘ 0“ ’3‘ -” '3’ -“ -.. U“
1 -5 -.. -11 .3. .as .1. .11 -00 -u no ‘ -1 -1o .11 -10 as -10 as 40 40 u»

 

 

 

 

 

 

 

 

 

 

no

no

no

1 E1 A 1
' CTT' I
I a 1 1 1 0 s a 1 a 1 ’ L 1 1 4 s s 1 a a
CVV‘I‘I'r
'4 L 1 1 1 1 4 1 a 1] l a 1 1 1 1 s c 1 1 1
3 4L 4 1 j 0 -14 1 I 1 ’ o 1 -1 1 s s 1 1
w- v' j 't' cows
2 o 1 1 1 a s a 1 a 1 ' 1 1 1 4 1 0 1 a 1
‘ a 1 .2 a 4 1 0 1 I 1 ’ a 1 A 1 1 a a
tTVI'YIII TIWI: 'WILI‘U Tirvwr "0"
1a 1 L111 1 1111 1 1111 o cm 1' i o 1 1 1 0 1 1 1 1 1
a A a a 1 a 1 ’ a 1 s s j a a
r LY

 

 

 

 

 

 

 

 

 

 

 

 

0 -1 1a 16 11. 0 -1 -1a - - 30
'1STITE Phil IS! ‘. NATIONAL PARK USE
I .l 0.. up -¥ .“ . u n o o -
H‘ . I I J J x
"Uri! 71 7"“ l x
Hunt:
1.1.? County 51111-3

O,'t‘UPATIONI

 

133

 

 

 

 

 

 

Iorurlrxcar1oa nun-Ia

, 0 1 1 1 0 s s 1 a I o 1 a 1 1 s
1 a 1 1 1 0 s s 1 a 1 1 L 1 1 1 a a
1 a 1 L a a 1 L 1 4 a L x

slcursrslus "nitric
A-Ol‘ m .‘03‘ m

11cn1cs1ac LOOIIIs
tug)” n. 0"." '-

sulnnxuc vans-16
51.05. '8 P-DS‘ '-

suunnrnxlc Infra 51111116
6-01” m I-O.‘ 1.

51111 01111115 nororlonrxuc
l-OQ‘ us J-Ion m

31111.11”; cauotxuc
l-lln "3 Plan I.

aouaoarxuc soar rxsuxuc
n-ulD VII "Na na

aaar rxsnxnc rxsuxlls-lanlus
04,1111 m you-a ran

sans; rIAIL 111111116
0-17. m I-IO. '-

uoaszaaca 1101-6 ”1"l'l'6
3.1.1» us run us

10011: vxslrxlc
u-n” I. v-um m

IILAIXIIG PHOTOGRAPH!
*2)” II. ].

orura TOTAbuACTTVT‘I'TII
7"“ 1 J L J— L l

ac'rxvzrr
1 o 1 1 1 a 1 a 1 a 1

nouns ACTIVITY

1 O I. 3
’ nouns

 

 

1 A L .1‘. J L L L I l L
A 1' V TY
, . . s . .
HOURS I
Ill J__I_m 1 2111 l 3111 C 4113 5.

 

APPENDIX E

TYPICAL "EXPERIMENTAL" SAMPLE PARK

TEST INSTRUCTIONS

134

.cEaHoo awn» aw
mvmm uxw: no sumo choowm

.z.m oouoa um
anonsumm can .z.m oouoa
um mmowum mmxon uwmommv
Eonm momma uumHHoo

.xon

.voumzmsxm ma aammsm us»
Hausa mmvuaumm can hmcwum
Ham mcumo #90 can: can
amkum .z.< oouw um :fimwm I
"maaconom vacuum: was» onHOM
wmmoam .Homm ma Hmnummz map NH
.vmumsmsxm mw aflmmzm
Hausa use can: can Amumm
ow. momma com saws unmum
.z.m coum um canon Gamma u
.voumzmaxm ma aammam
may Hausa age can: can
Amuom oma. mcuoo com :ua3

uumum .z.< ooum um gamma u .umm
.cmumamcxm mw xammsm
may Hausa use can: can
.muom on. mouse oma nufis
uumum .z.m oon~ an gamma a
.voumzmaxm ma yammsm
ms» ”was: use can: can
..muom omv moumo omH nu“:
uumum .z.< ooua um canon . .flum

no : osom unoccm: mwsu soaaom

mmmmam .000 mg Monumm3 0:» NH
.mufimccowumwsv no» man :ufis

wcficcwmmn Hoouo cw use Ems» can:

 

Ammqmz¢xm7mmmv
ammoqozm

.3OHHM

ocwuuso 0:0 can
xouHa mco nu“;
uumno a m“ mm
coflumusoumumuu

.msonum xomHn

n ma: nods:
HHQSU 5 fl“ MN
:ofiumosouenmu~

n.0wc mowuw>wuom

-cfiwun mw :ofiu
luosv mmnumnmum

”.muwuw>wuom 0:»
gums. nu“; mafia
nun vN counummaoj
.vw can mm mcofiu
omwsw mmzumumuw

:..3Oamn vmcwp
noun muommm ma»

gowns. an“: mcfi=<

o~ wash

:H: sufi3 mcwm.>mcusumm

 

 

:ofiuomaaoo 0:» GM oouwmom .umnummou cmama awn cm coHumoso can
two Hanssc mg» mam8«m ma nHoo mcwmuam so» on omwaamsm ma Hnfl3 .vw can mm mcofip ma mash
vacusuwu umnfiac was Wfiuwmccofiummav ucmumumwc 0>wm may Immsw manuauanm amoflum

p30 amnz<x mzc~huamkmz_ mm~<zzo~hmmnc mk<o

Guzmzhmm mmmz:z

 

 

 

 

 

umfiwuuwm 23~H<uoq 02~Hmmh

 

mzum axqusmxquzo_bmmsa Eqbzuzwxmmxm . >aapm mm: >qa

v¢ mumguoum‘wmvon xm<m

135

 

 

mIMIN

 

 

 

.w>wm Amv
vumnm mo mama
mo ummmmc
mnmum
NIMIN Too Qmoz<m 802
HnH:~ nommm:
adamafidm
can ooum .ucmwcm>:oo muos ma cHso3 u“ .mconm an»
minim Hummus umm: ummx 0n canon nowumEH0waw mfinu
HH .umaamo on» o» cucusumu uwnfidc
vums~ can use caucus Hogan: msu ommu mummam
can ooum .wmocnm no cacosm on Haw; now
momnm wéamsacm .mcfimcmg
.muflmmo aowmfi>fio 0» sun» uumzuom
Namum memmc can .805» Hman .mcumo no maccaa comm
caucus momma owummam usm mmmwam
H-H-~ .mcfiumm: guso
can ooum conga; «on: map umnc: umnasc mnu umuzo
rdonm H0 2 can use cmccmn no: Hawum n no mumm
may ucaoo .mmuflmcco«ummaw mo umnEd:
cmcusumm Anna» 0:» use new: on manwmmanfi ma
can ooum ya “a .um>mzom .wmvnsumm can amcwum
wmaHmm :0 momma Ham use can: on umsmuud
F30 amaz<z mZGHPUDmhm2_ wm_<zzo_hwm:a mh<a
Guzmahma mmmzaz

{It

 

 

 

 

 

mznm amquuquzzofimgo .ECJEEEE - >35 mm: >65

Mocmuucm zc~H<UOJ oznbmmk

 

¢* mumsuoum mmooo xm<m

APPENDIX F

DESCRIPTION OF VARIOUS

"EXPERIMENTAL" QUE STIONNAIRE S

APPENDIX F

DESCRIPTION OF VARIOUS ”EXPERIMENTAL" QUESTIONNAIRES

 

 

 

 

 

 

Basis for Paper Questionnaire
Test Design of Identification Comments
. . Stock
Questionnaires Numbers
1 Same content Blue, 1-1-1 Folded card
and design of 24 1b. , with slotted
”seasonal" Carnival hole in upper
questionnaire Kraft left corner.
Blue, 1—1-2 Folded card
24 1b. , with no hole.
Carnival
Kraft
Blue 1—1-3 Folded card
Index with slotted
Card hole in upper
left corner.
Blue 1-1-4 Folded card
Index with no hole.
Card

 

 

 

 

 

136

137

 

 

 

 

 

 

Basis for Paper Questionnaire
Test Design of Identification Comments
. . Stock
Questionnaires Numbers
1 Same content Blue, 1-2-1 8% X 11 inch card
and design of 24 1b. , with slotted hole
"seasonal” Carnival in upper left
questionnaire Kraft corner.
Blue, 1—2-2 8% x 11 inch card
24 lb. , with no hole.
Carnival
Kraft
Blue 1-2-3 8% x 11 inch card
Index with slotted hole
Card in upper left
corner.
Blue 1—2-4 8% X 11 inch card
Index with no hole.
Card

 

 

 

 

 

Due to the success of the folded card and slotted hole, all the follow-
ing questionnaires were folded and contained a slotted hole unless
otherwise mentioned.

 

2

 

Basic design
from the
”seasonaP'
questionnaire
with the
exception of
the activity
analysis
questions
(questions 23
and 24)

 

Yellow,
24 lb. ,
Carnival
Kraft
used
through-
out test

2-1-1

Same activity

analysis as in
the "seasonal"
questionnaire.

 

2-3-2

 

2-3-3

 

2-3-4

 

 

2-3-5

 

See Figure 11,
pages 141-142
for the activity
analysis ques-
tions used in
this test.

 

138

 

 

 

 

 

 

 

 

 

 

 

Basis for Pa er Questionnaire
Test Design of p Identification Comments
. . Stock
Questionnaires Numbers
3 Same content Yellow, 2-1-1W "Seasonal" ques-
and design of 24 lb. , tionnaire with a
"seasonal" Carnival pencil enclosed
questionnaire Kraft in slots cut in
used the bottom of the
through— front page.
out test
2-1-1N "Seasonal" ques-
tionnaire with no
pencil.
4 Same content 24 lb. ,
and design of Carnival
questionnaire Kraft
2-3—5 used
through-
out test
White 3-1-1
Yellow 3-1-2
Blue 3-1—3
Orange 3-1-4
Brown 3-1-5
Tan 3 - 1 - 6

 

 

 

 

 

139

 

 

 

 

 

 

 

Basis for Paper Questionnaire
Test Design of Identification Comments
. . Stock
Questionnaires Numbers
5 Questionnaire White 4-1-1 Exact copy of
2-3-5 was Index 2—3-5.
used as Card
content and used 2-4-1 Post card types
design basis through- 2-4-2 with "deposit"
out test 2—4-3 message on back.
See Figures 12
and 13, pages
143-144.
4-1-2 Similar to
2-3-5 but with
no slotted hole.
2-5-1 Post card types
2-5-2 with ”mail back"
2-5—3 design on back.
See Figures 12
and 14, pages
143 and 144.
6 Questionnaire Yellow, 5-1—1 Exact copy of
2-3-5 was 241b., 2-3-5.
used as Carnival
content and Kraft 5-1-2 Same content and
design basis used general design
through- as 2-3—5, but
out test with questions
arranged in
different order.
5-1-3 Same content and

 

 

 

 

 

 

general design
as 2-3—5, but
with questions
arranged in a
different order
from 5-1-2 and
2—3-5.

 

140

 

 

 

 

 

Basis for Paper Questionnaire
Test Design of Identification Comments
. . Stock
Questionnaires Numbers
7 Yellow, 5-1—1 Exact copy of
24 lb. , 2-3-5.
Carnival
Kraft
used 2-1-1 ' Exact copy of
through- "seasonal"
out test questionnaire.
8 Exact copies White 2-5-1 Previously
of question- Index 2-5—2 described.
naires in Card 2-5-3
Test 5 used
through-

 

 

out test

 

 

 

141

5.63m won wagoflmosmv

mo owmm Mommuum... ”Emu. Hmpoofiwaomxm E mofimccofimmsa “no...“ mooBmosG mambmfiw Nan—“>304: .3 mmDUHm

m n m u N on Race Sm 3.0

N: m u m on .mfifioflwosG

H .. H u m on EncoflmosG

 

8...: ...-no! 63-5... .35
33.2.5. 83m 555.:

033.3. 8883: 33.2 .3... 88».
...... 3.5:: ...: €338. 8:333.

.3 = :15 0301 Jul. 05 Bob F)- E :5 vein-o b.8300- obqn 3.. a

$5: 25 Him 92 5.252“. 5. w><z Ed: m8» mo“. 82$:

3...: Help! 6333.. .5
3.323.; g 535....

3.2.3 82.68: 7.53.. .8» 58m
:5 9553:. .3. non!!— con-88¢

33:13:81— JiagunII-E-avogobaggzfliu

98: .5: 93m $3 535sz .3 w><= Ed... ”.2; «on. 8.25:.

8...... 5.822! £533 .3“.
3.35.9. 85m 535.:

5:5 22:23. .235. .8». ES...
:5 2:53.... ...: €233. 8:353.

.3 = ....E 0301 Jul. 05 59c .33.. End 35 var—.30 .3380- !in no» a

“to: “:5 E5 mz< “1:25.15 5 m><x Ed: ~59 «on. 82$...

 

 

 

 

 

 

 

 

 

 

9:5... v

P59. 0

953 A

P50... I
flag 3:3.3.

.VHZLP—hyqt. NI... ho =9<N 209 PZHLm .uUOzO ¢DO> Mun—.3:
LO ZNQEPI. ...—:F 7.“ HHVS» O.rJ< .CZEOG mink PW»: ...:C. Hilde—v. QDCKU
EEC? WEt.C.—.b< mph: m2... .6 “wk—K3 .BOJEE Quararim manna...“ NI... a QN

 

 

 

 

 

 

 

 

 

 

551333

3!... 25... 3953

«an a a ”15:3 98: 8 EDOKO “DO? SO EFU< 22:? MN

 

local 1 US 52. nulllué
BB. I _U ... Baal -- l D ...
«8 Iggy—D HIP an #35 '8
In} 1. D .o :34 1.1.1.11... 33 D .0
{all 2 D .n 953' ......... 3.55.5 D ...
3331'! n- D .c 950.. ....... 1.05.33: D .u
aim 3 males 3%
non «.539. ”...... 89 CE? .52.. 5:. z. adv; :uFruch
.Ehtsg Emits. on Bo E520 uhjazoo 5.0»
o... .3 Sum 5.98 .30» b #5: EPEU... mu... 5?..- VN
.5 375.
.350
«5555.. D .u
33.9804. D ... .
Annals. uEz-u D a a. D .—

5332 D... .32.... 45... D .o

.33... :3 D .c

3312!?” D .6 .309. n .5.
£23 ..a 2555.

8 BE - .0
E35... 9.35.. 5 F D

.3 3563:. 95.2.. D ...

953 3.3.. 233 D...

a: 00.... .
5.3 . u D _ .55... i .398. D .u
..‘E 8 355...: D .. as...» D ... 9.22:0... D ...
.51: 1933.34. D .a 85.-€038 D A 35.x.
212.2...D; .33... 2:8. o. 2.3:. D A
Esau 1:9. .3 sin D ._ 3:0 .23
Eco. v5 ...-Elm D ... :5!- ..3c! D .n 596 5:905:5- D .-

..:fi. :5 .38.. on. ... .86. a, 3......
3...... on 55...... 5.0» Ba 33mm swam: WEE? 5F .5 5:5 MN

...: 3:3.

 

.35».
...—3.35.... D .u
Anna-i. union D .n 8:39! B ..
~55... :5: D .o 3555;. D .-
use... :8 D ...
act-3.30.. D .E .32.... u .5.

. 5:2 ... 158.5
3.023.. D _ .55... .... ecu—:5. D .v

9.3.5835. D .x
U552?— D. .-

23:9... ..... an:
.3 $5.53.: “3.55; D . I

9:5. 322. 95.3 D ...
3...:

1135....6. .2... D .. 5...... D ..
use... 19309.2. D .o man-.8508... D ... 350...
9.2:. 2!: D ... an. ..a. Econ. 0. 95:5. D .3
Bi: 3:0... 50 :0: D ._ Eco ado
=33 va- nes}... D .0 :5!- »81.’ D .a SP: 3305...»... D .-

...a... as .33 as :- 126. $5.”...
5:...» on 5.9.0 58> Eo BS... out”... REFEU.‘ “:F .3 5:5 MN

 

 

 

 

 

 

.A

 

.52.. a..." 8 939. H... as 6.... as... 2.303 fink—do:
:0 Ease...» 22:58”... ..<zc:..ac< ...o 98. as...» NN

 

 

 

.0

 

.n

 

3:2,. 5; on .59? HF. E 5.... .3.» 5:3... mustard:
:0 E.E:Eac 22:58.“... ..<zc.....cc< ..5 8.5. 52.3 «N

 

 

 

.u

 

.0

 

.52.. 9F 2 955 man 2. 5.... so.» 9.503 25...:
so EzEkOanc fixing"... .....zCEaE .3 88. as...» NN

 

 

142

9353.80 nu .HH HmDUHh

mu m u w my $58 .5650

T m u N madame Emmsg

 

 

88v 8340! 653... .9...
3.9.3:: 8.3 5532:

3.2.3 828.8. .58.: £8 ES:
...... 3355 3.. 5288: 8.89.8:

"8:55.301 Jaaghch-avotaobugggaih

g: ...:E wnam $2 5m<>owzm z< m><= Ed: «8» mo”. $.25:

 

 

 

i ll gugua

 

«3553:...

 

—:gummm— «IN .N .Im ":13

 

lulu-om ..80 .953

 

a... «333.8489

 

Salami 5.52 £8 5

 

 

3.. «al.- 05 no .525

05 5 8th 030... .08-

 

oaorm c.831— 480 fig

 

 

 

 

 

 

 

 

 

 

 

 

 

O Q O «98.8.83.

 

0559”.“ hams-honflofl
gnu-fio< 55739. oflzlfo£58t388n MN

 

 

 

 

 

.0

 

.n

 

3:7... BE. 2. 859. mum 2 8.... .5» 5:03 359:.
no Eamoaao 229:5”... q<zgnn< no 58. as; «N

 

 

v.23 :58! 6533 in
9:83: .88 8242:

8.2.8 883.3. 15.2 .«8 68m
=5 .53.... 3- .288: 8388:

"83:15:31. gagingigghiggflzl

£0... .5: gm 92 59.53 5 w><= Ed: 55> «on. 8.25:

 

           

  

'"DMM 9W
)0 w In und- bm mm
110nm nan moqn on.“ mold

  

' 339 z 1- Iznoq

3.58839
asgué3gfiua

Signing 33NMM§S8E§EGN

 

 

.0

 

.&

 

«gm BE 2 Gun—Q4 awn OF Hun—A 30k 383 §§O<h
GO Ear—2A0 ZE<§U§ A<EGQ< hO DE PE NN

 

 

 

143

3* cam 3* Emma. Hmpcmaflmmxm E ummb mmfimfihoflmosG :vpmo pmom: mo 33> xomm :1 .NH mmDUHm

mumum cam muvum mmuwmgoflmmng

m-m-m 28 Néé 3:438:85

 

 

$5: .5: gm 82 5352”,. z< u><z Ed: ~52 «on. 9.2:...
38...).

Ian—U;
Diana-Ill-
ibhln-o: IBIBE'ED...
via-3:06 i

9.8 ital-3&0...
gliiud gagi'ad
alzsflaflnvgle .Eighoghe .

 

.a... 83:2: U .v
.Bd SKI-Boa D .o
.3... 86.86 D a 88.8"... D ...
.8... 86.86 D .. .ai 8.2.3.. B .-
38 :88. 3.5:. Hz... muta— aor an HE. .523 n

...—d 8678... D .-

 

..:- uuuuuuuuuuuuuuu

«H30:
‘8? 30¢.— gn— IE a .HhDOm PURE—D h; HIE rm .3 g 30: ‘

 

 

A.><3 Hut. 23< a 2.359289. 29.340!— FOE
wh<8fi at U: 0.... 944m><¢h Ola-Am 90» an flak IRS in! n

 

................. 9 “Eco Eu .50» a 955 N

 

85m 5550 Dru .8 nick

69:32 Us .923: .09:- 235 F 8.0: ~22qu .50» fl nuns? —

 

zoCmmso >aw>m .So .3: um<ma€
50......ch SIS £33! .3 can-cahoou ”Sad-E E 9.58
PLOP—Oreo 05 33 ESICUaoU .5 EuEanlH ...5 Lo. 39.580 8.3 3 .€3o IE...
..Squauvcgu “.9275 on :t-
.39.... no» :5 Sada—ESE :n .0230... *0 .vs and-Unhng 5.3.9.5: «A El 30: 25>
mutizt no m4aaou < >420 mug. 44 _l : 23.83 2823.6 :- :8 :c 2.105
3...! 30> mention
I: .3593 v:- ESP? :3 .5: 9.5.3:: 2!. 8 v00: on :7 9.3.. 9:.— .5.» sonar—£5
.5. ing 38 use on: oat 9.9.99: 05 30.0.3... 3 c.8920. 035 .o 2! 93 30>

5. a. 3., +351!!! 5 E is.

 

Humum cam Huwum mmfimzcoflmmna

 

 

 

 

was: .5: gm oz< Wage—.5 z< w><z Ed: «8» mo... 025:

5.8.9.8.: D .u
:8 v8 8...»: Di 08.5.89... Dd
.8.§.8o.fl. D; «8.8.98.8 D ...
8...;6863 D.- Sod .85. D .-
..8 486. 5...: $5 a; 458.»
SF Baum—a tum 23349. a: .5 22:3 o

0.35.82 3 2 I 2

DDUDD

3 z 2 a ... p o a v n

D a DD Daaanmm

 

 

HAD—rm.»
Col-a- eao guano. s. ..>.:l<m 55> Lo Q<m=: EC. rm
ant—.5200 th<uznm ...O 33”.; mo mmmipz 1399—. NE. LO
5:. “£2952. ham 3C4“; 95.39:. 5: L0 :oEB m
3. SF: 8:359: ass
HE E
... 05.5
< :0.— 8 ..».-l<u Know to 35... 5: Eco .535 v 93.33!
.2... D _
h 5; Ice. .......... ... “9.. an 2.
....E D .
r £131.; .30.» no 35:...
HE no m0< whixcxnn< P27. Er. 5.5. v.— h<:.3 n
. . 33¢.sz
.1'I .III .I .II ..III III nwddxx

33.. 4._:..._> 7.2:. .5 “.4.—Qua 5:. h: WMU< HIP m5 F; N

 

.... D t... D .EEE 7.5 :2 ......E a as: .5» Sn —

 

3.0:mwso ..umsh ~30 .3: wm<uqm
.C_E........_.. 35m 535...). .3 5:55:80 25:55 no 0080
Pugh—950 05 v5 cozstoréFu .c «8.5.:«3: 95 L5 V3259! 5:3 2 Gan 3FH
3:19—96:80 385 on :«i
.3995 :9». :5 53:62:. :1. .3300 .3 £5. 18598,“? ’33.; ...1 ...! 23 ....oh
mwbbz; uo wJADQU < >470 w¢<h 11:1 .5 3539:; 155.339 :q =6 :c 0:65
so?! no.» .33»on
.5 03.5.3 33 53:»... 1:: =5 95.95 Eve 3 7.": an :7 ES. 633 7?». Samar—35

ofl. if... 33w. ...: 23 to: £38 £3 E32?! 3 39:... .30.: go 03 9..- so.»

 

. 5.18» 33!... gg.§fl=§i§

 

 

 

5 55.— g 82 3953 .2 w><= Ed: :8» me $25:

 

 

 

 

656-6

H“

g ”a

go< "a ho ID: Ego Agog;
hoggagg .ESFEEEHAna—DOQO
flsrggfllhl—E .38: 8990.: nus—DHIFE N

 

E. 8:).

 

has
0581' D .-
RfluIDi .
I231! D o 3!!)- D ..
38.023: .a
35530.» .5333 D
389.2 . 348.8.
pinion-Inna... D...“ .25 .8 .155
13 . :2: D w .aia..ua18_ D ...
iii—383.0 4 3:1. D J 50.8... D .0
35:13:80.. 23%... D .. 858
Indiana; .52.. 3823231.. U ...

Eon- 350- no 51.. D é £8 .39
gal-ED... gal-5:0... Egg-D .n
7383305213 pg
".5 8 “92° “8% an 38mm ONES §U< HE no In: —

 

3:330 >¢w>w .50 .3: um<u4m
.bisis 33a 533.: B 8:38.89 aqua-E B 880
...—8.950 In .2: con-£280 3 «8:5an .5 5. v3.9.9.8 ~58 : 33. a
£538.38 v88: 2 5..
b3:- 2; is 855.55 :1 .9958 .o .3. 3:98....- haqoh. 1. :F 38 :6»
35...: mo 3.38 < 55 8:: j ... I .2323 353.... :- .8 :c 3!:
snarl so» suntan-

JS 8.39.“ v... 88-»- nul .28 25.3.5 23 3 no... on :3 80: 2:5 .5» 8:15.82:
a .355 818 .56 0!. of. 2.8!. 04. 30039.... a. 0.3020- 935 .0 98 8c 3%

‘ { TIA 1' ”11". I I

 

 

 

144

 

Plflii III'IISII
IHI Illi I'll! [III

III: "III IISI filll
II "II lllll

 

 

 

FIGURE 13. -- Front View of "Post Card” Questionnaires (2-4-1,
2-4-2, 2-4-3) Used in Experimental Test #5

 

No
Postage Stamp

Postage
Will be Paid
Neceuary
II Mailed in the
United State:

by
Addressee

   

 

BUSINESS REPLY MAIL

FIRST CLASS PERMIT NO. 698. LANSING. MICHIGAN

 

 

 

DEPARTMENT 0F CONSERVATION

Recreation Resource Planning Division

LANSING. MICHIGAN 48926

 

 

 

FIGURE 14. -- Front View of "Post Card” Questionnaires (2-5-1,
2-5-2, 2-5-3) Used in Experimental Tests #5 and #8

APPENDIX G

”EXPERIMENTAL" QUESTIONNAIRES
HANDED OUT, RETURNED,

AND COMPLE TED

APPENDIX G

"EXPERIMENTAL" QUESTIONNAIRES HANDED OUT,

RE TURNED , AND COMPLE TED

 

 

 

 

 

Test Questionnaire Number Number Number*
Identification Handed Out Returned Completed
1 1-1-1 908 435 435
1-1-2 250 102 102
1-1-3 579 278 278
1-1-4 579 285 285
1-2-1 579 278 278
1-2-2 579 250 250
1-2—3 579 305 305
1-2-4 579 286 286
2 2-1-1 356 146 84
2-3-2 358 152 99
2-3-3 356 138 83
2-3-4 357 142 79
2-3-5 357 161 92
3 2-1-1W 500 240 240
2-1-1N 500 163 163

 

 

 

 

 

considered to be complete.

>I‘In Tests 1, 3, and 4 all returned questionnaires were

In the remaining Tests, all returned

questionnaires were individually examined for completeness of
response.

146

 

 

 

 

 

 

 

Test Que stionnaire Number Number Number
Identification Handed Out Returned Completed
4 3-1-1 274 158 158
3-1—2 273 147 147
3-1-3 273 159 159
3-1-4 273 141 141
3-1-5 273 157 157
3-1-6 273 153 153
5 4-1-1 387 124 102
2-4-1 387 114 93
2-4-2 387 166 138
2-4-3 387 112 98
4-1-2 387 119 97
2—5-1 387 104 80
2-5-2 387 122 102
2-5-3 387 137 128
6 5-1-1 396 168 128
5-1-2 396 156 122
5—1-3 396 172 123
7 5-1-1 500 171 123
2-1-1 500 178 126
8 2—5-1 500 71 66
2-5-2 500 74 74
2-5-3 500 86 86

 

 

 

 

 

HICHI RN S

III/HIM!!! mil/I

Iii! III/I

406

iIIIIIM L ““

31293104 I'll W

I