Up to now you have learned about specific thinking and some action processes, such as the role of theory in research (Chapter 6), ways to identify and frame research problems (Chapter 4), and the purpose of and methods for conducting the literature review (Chapter 5). All these processes are part of the important work of refining the structure and content of any type of inquiry. Now you are ready to move beyond these initial starting blocks as we examine how to develop specific questions within the experimental-type tradition and queries within the naturalistic tradition. Developing a question or query represents the first formal point of entry into a study.

We have just used two distinct terms, “question” and “query,” to describe these initial formal points of entry into a study. These terms reflect the different approaches to research used in each tradition. Your research question or query will guide all other subsequent steps and decisions, such as how you collect information and what other types of procedures you will follow.

Research questions in experimental-type design

In experimental-type research, entry into an investigation requires the formulation of a very specific question with a prescribed structure. That is, from a broad topic of interest to you, you must articulate a specific question that will guide the conduct of the investigation. This question must be concise and narrow and must establish the boundaries or limits as to what concepts, individuals, or phenomena will be examined in the study. The question also must be posed a priori, or before engaging in the research process. The question becomes the foundation or basis from which all subsequent research action processes are developed and implemented and on which the rigor of the design is determined.

The purpose of the research question in the experimental-type tradition is to articulate not only the concepts, but also the structure and scope of the concepts that will be studied. The research question leads the investigator to highly specified design and data collection action processes that involve a form of observation or measurement to answer the question. Research questions are developed deductively from the theoretical principles that exist and are presented in the literature.

Each type of question reflects a different level of knowledge and theory development concerning the topic of interest. Questions that seek to describe a phenomenon, referred to as Level 1 questions, are asked when little to nothing is empirically known about the topic. Level 2 questions explore relationships among phenomena and are asked when descriptive knowledge is known but relationships are not yet understood. Level 3 questions test existing theory or models and are asked when a substantial body of knowledge is known and well-defined theory is developed.

Level 1: Questions That Seek to Describe Phenomena

Questions that aim to describe phenomena are referred to as Level 1 questions. Level 1 questions are descriptive questions designed to elicit descriptions of a single topic or a single population about which there is theoretical or conceptual material in the literature but little or no empirical knowledge. Level 1 questions lead to exploratory action processes with the intent to describe an identified phenomenon.

You are a therapist working in a rehabilitation center. One of your responsibilities is to introduce patients to a range of adaptive equipment that they may need or may find useful in the hospital and on their return home. You notice, however, that some of your older patients appear hesitant to use this equipment, and you are not convinced that they continue to use the issued assistive devices once they are home. Because this has important practice, policy, and cost implications, you and your department consider this concern important enough to investigate.

Your first step is to conduct a comprehensive literature review. You might start with a keyword search, such as “adaptive equipment,” “assistive devices,” and “rehabilitation.” The literature review shows only a limited body of knowledge regarding adaptive equipment use and little theoretical development to explain patterns of use, particularly among elders discharged from rehabilitation. Thus, on the basis of this review, you realize that the first step is to design a Level 1 study to describe patterns of equipment use in the home after rehabilitation and identify reasons for use and nonuse among clinical populations, such as older people with recent strokes or hip fractures, who would benefit from the use of equipment over time.1 Your initial Level 1 research questions may be as follows:

What is the frequency of use of adaptive devices by older individuals who have been diagnosed with strokes and discharged to their home after rehabilitation?

What are the reasons for use and nonuse of assistive devices by this clinical population at 3, 6, and 9 months after discharge?

For example, a body of knowledge exists regarding the attitudes of occupational therapy faculty toward including students with disabilities in occupational therapy education. However, little is known about the attitudes of other public health faculty toward including these students in graduate education. Therefore, research based on a Level 1 question will be meaningful and appropriate to describe the attitudes of public health faculty toward accepting and teaching students with diverse disabilities in their graduate classes. Level 1 questions focus on a description of one concept or variable in a population. To describe a variable, which can be defined as a characteristic or phenomenon that has more than one value, you must first derive a lexical definition of the concept from the literature and operationalize or define it in such a way as to permit its measurement. Thus, Level 1 questions focus on measuring the nature of a particular phenomenon in the population of interest.

In response to the changing context of practice, a recent study conducted by Hoppes and Hellman² examined the attitudes of occupational therapy students toward community practice. Although there was a growing body of literature on the need to educate students for community practice, Hoppes and Hellman were concerned that “students’ voices have seldom been heard in the discussion about assimilating community engagement into occupational therapy education”. To inform progressive occupational therapy education that would be meaningful to students, these investigators initiated a descriptive study to answer the following research questions:

Do students believe there are specific actions they can take to address community needs?

Do students feel a moral obligation to help in their communities?

How do students perceive the costs and benefits of helping in their communities?

To what extent are students aware that needs exist in their communities?

To answer these questions, the investigators identified specific items on the Community Service Attitudes Survey that measured each of the variables names in the four Level 1 questions. Occupational therapy students were used as their population of interest. Box 7-1 presents additional examples of Level 1 questions. Note that they answered the questions “what?” and “how much?”

BOX 7-1

Level 1: Descriptive Research Questions

1. What are the most commonly applied for assistive devices in an elder population of Swedish residents who apply for grants for these devices?

2. For elders with depressive symptoms, do impairment, functional status, and severity of depressive symptoms change over 3 years?3

3. What is the incidence of a history of family problems in the area of substance use, psychopathology, compulsive disorders, and violence? Are there significant differences in student histories of family problems by age, gender, or race/ethnicity?4

Think about the basic, descriptive research questions in the previous examples. What is the main concept or variable that will need to be measured in Question 1? It is the most common adaptive equipment being requested. What is the primary population that will need to be recruited? It is elders in Sweden applying for grants for assistive devices. Question 2 seeks to describe the variable of functional status over time, and Question 3 measures the variable of family problems, defined as substance use, psychopathology, compulsive disorders, and violence in a population of social work students.

As you can see, the focus of a Level 1 question is on “the what” and “how much.” Level 1 questions focus on the description of one concept or variable in a population. A variable can be defined as a characteristic or phenomenon that has more than one value. To describe a variable, you must first derive a lexical definition of the concept (description of the term in words) from the literature and then operationalize it (define a concept by how it will be measured). As an example, in the study by Sellers and Hunter,⁴ the lexical definition of family problems is defined by the presence of substance use, psychopathology, compulsive disorders, and violence in one’s family of origin or childhood.

Level 1 questions focus on measuring the nature of a particular phenomenon in the population of interest. These types of questions use the stem of “what are” or “what is” and refer to one population. As shown in the examples, only one population is identified, such as elders applying for financial support in Sweden, elders with depressive symptoms, and social work students. Also, single concepts are examined in each study, such as assistive devices, functional status, and family disorders. Each of these concepts can be defined in such a way to permit their measurement; that is, they can be examined empirically.

Level 1 questions describe the parts of the whole. Remember, the underlying thinking process for experimental-type research is to learn about a topic by examining its parts and their relationships. Level 1 questioning is the foundation for clarifying the parts and their specific nature. In the scheme of levels of abstraction, Level 1 questions target the lowest levels of abstraction: concepts and constructs (see Chapter 6).

As discussed in later chapters, Level 1 questions lead to the development of descriptive designs, such as surveys, exploratory or descriptive studies, trend designs, feasibility studies, need assessments, and case studies.

Level 2: Questions That Explore Relationships Among Phenomena

The Level 2 relational questions build on and refine the results of Level 1 studies. Once a “part” of a phenomenon is described and there is existing knowledge about it in the context of a particular population, the experimental-type researcher may pose questions that are relational. Level 2 reflects relational questions and builds on and refines the results of Level 1 studies. The key purpose of Level 2 type of questioning is to explore relationships among phenomena that have already been identified and described. Here the stem question asks, “What is the relationship?” or a variation of this (e.g., “association”), and the topic contains two or more concepts or variables. “What is the relationship between exercise capacity and cardiovascular health in middle-aged men?” In this case, the two identified variables that are measured are exercise capacity and cardiovascular health. The specific population is middle-aged men. As you can surmise, Level 1 research must have been accomplished for the two variables to be defined and operationalized.

You are involved in establishing exercise programs for adults as part of their cardiac care. Although you know there is much evidence to support the cardiovascular benefits of aerobic exercise in general, you do not know much about the exercise capacity of middle-aged men, and this group has become your primary clinical population. Thus, you may be interested in asking a Level 2 question such as, “What is the relationship between exercise capacity and cardiovascular health in middle-aged men?” In this case, the two identified variables that are measured are exercise capacity and cardiovascular health. The specific population is middle-aged men. As you can surmise, Level 1 research must have been accomplished for the two variables to be identified, defined, and operationalized.

Returning to the example of attitudes of public health faculty, suppose we now want to know the relationship between attitudes and numbers of students admitted into public health majors. We would pose a Level 2 question such as, “What is the relationship between faculty attitudes and number of public health majors with mobility impairments. This Level 2 question would lead us to measure and look at the association between two variables, attitudes and number of majors admitted to public health programs. Suppose we found that negative attitudes were related to low numbers of public health majors with mobility impairments. We would be able to claim an association but not a causal relationship between the two variables.

Let us revisit the Level 1 examples and see how they can be modified to become Level 2 questions. Suppose we have conducted studies to address the Level 1 questions in Box 7-1. If we continue our research agenda in each of these areas of inquiry, we would be ready for the relational questions in Box 7-2.

BOX 7-2

Level 2: Relational-Type Questions

1. What is the relationship among difficulty and independence in performance of activities of daily living, use of assistive devices, housing, and living situations to the home modification applied for in a sample of applicants for home modification grants in Sweden?

2. What is the relationship among depressive symptoms, impairment status, functional status, and use of assistive devices?

3. Are there differences in student histories of family problems by age, gender, or race/ethnicity?

Level 2 questions address relationships between variables. Studies with this level of questioning represent the next level of complexity above Level 1 questions. These questions continue to build on knowledge in the experimental-type framework by examining their parts, their relationships, and the nature and direction of these relationships. Level 2 questions primarily lead to research that uses passive observation design, as discussed later in the text. Refer to the levels of abstraction in Chapter 6 and see where Level 2 questions fit in the schema of building knowledge in the experimental tradition.

Level 3: Questions That Test Knowledge

Level 3 questioning builds on the knowledge generated from research conducted in Level 1 and Level 2 investigations. A Level 3 question asks about a cause-and-effect relationship among two or more variables, with the specific purpose of testing knowledge or the theory behind the knowledge. We refer to this as a predictive question. Given the findings of the Level 2 question, it would be important to know whether attitudes are causal of barriers to public health education for students with mobility impairments. A Level 3 question would therefore be indicated. Building on the knowledge already generated, we now would be able to ask the following question: “To what extent do faculty attitudes influence the number of students with mobility impairments who are accepted into public health majors?” The resultant study will test the theory behind the reasons that negative attitudes create educational barriers, and the study will be able to predict the opportunity for these students to be accepted into public health majors on the basis of faculty attitudes. At this level of questioning, the purpose is to predict what will happen and provide a theory to explain the reason(s). On the basis of a Level 3 question, specific predictive hypotheses, statements predicting the outcome of one variable on the basis of knowing another, are formulated. Action based on this knowledge can be taken to promote educational opportunity if you know where, why, and how to intervene.

At Level 2, we asked the question, “What is the relationship among the attitudes of occupational therapy faculty and the number of students with disabilities admitted into occupational therapy education programs?” Assume that in conducting this relational-type study, we find an association between faculty attitudes and number of students with disabilities. That is, our study shows that programs that have occupational therapy educators with positive attitudes have a higher enrollment of students with disabilities than programs with faculty having negative attitudes. Building on this knowledge, you can now ask a Level 3 question such as the following: “To what extent do faculty attitudes influence the number of students with disabilities who are accepted into occupational therapy programs?” Such a study would test the theory behind the reasons that positive attitudes promote opportunity for students with disabilities. Also, the study would be able to predict the opportunity for disabled students to be accepted into occupational therapy programs on the basis of faculty attitudes.

In a Level 3 question, it is already established that two concepts are related, based on previous research findings (from Level 2 research). The point of study at Level 3 is to test these concepts in action by manipulating one to affect the other. Level 3 is the most complex of experimental-type questioning. Once the foundation questions formulated at Levels 1 and 2 are answered, Level 3 questions can be posed and answered to develop knowledge—not only of parts and their relationships, but of how and why these parts interact to cause a particular outcome. Level 3 questions examine higher levels of abstraction, including principles, theories, and models.

Consider these examples from the two studies discussed earlier that contain Level 3 questions: