1. Introduction
Ecological momentary assessments (EMAs) are repeated samples of behavior in real time, most commonly used in health science research [
1,
2]. The approach is based on the theoretical foundations of behavioral medicine, which try to understand the relationships between social, psychological, and behavioral factors at the individual level [
1]. Traditional data collection methods in the social sciences (e.g., retrospective questionnaires) are subject to limitations in external validity, and uncertainties in how far the findings can be generalized beyond the investigation [
3]. Additionally, survey results are subject to recall bias and do not capture specific intra-subject variability, particularly when respondents have to generalize an experience or behavior to a certain timeframe [
4,
5]. Smartphone technology has largely taken the place of other devices, such as pagers or beepers, to collect real time data and remind a respondent to answer a question or phone the study coordinator [
6,
7]. However, any means of collecting data in real time or near real time is considered an EMA. To our knowledge, EMAs have not been used in forestry research to date.
We tested the EMA approach on family forest owners (FFOs) because families and individuals own over 40% of U.S. forestland and thus control the public goods and ecosystem services provided by these forests [
8]. These FFOs make stewardship and management decisions, which range in intensity from deciding to harvest timber from their woods, deciding to bird watch, or doing nothing in or with their woods. A long, rich history of scholarship has been devoted to understanding how and why family forest owners make these decisions (for a review, see [
9]). One conclusion from this body of research is that behavioral intentions do not always match actual behavior (e.g., [
10,
11]) possibly due to methodological and theoretical issues. Methodologically, measurement error and potential recall biases occur with self-report retrospective survey methods, which may impede measurement of actual behavior. Theoretically, the length of time between the setting of the intention and the expression of the behavior for most forest-related behaviors may make frameworks like the Theory of Planned Behavior [
12] a mismatch for the behavioral outcome of interest. Existing environmental behavior models may not adequately account for the relevance of forest management activities to a private woodland owner. Often, major forest management decisions (e.g., timber harvesting) are made infrequently, and involve complex issues. Exploratory research shows that landowners have high psychological distance, meaning forest management decisions are often abstract and lack detail [
13].
For infrequent decisions, understanding the entire decision-making context is important. This includes the antecedent internal cognitive factors that influence an individual’s decisions as well as the external social and contextual factors that may contribute to the decision [
14]. Internal and external factors in decision-making have been studied across a wide range of natural resource disciplines, with the general conclusion that both must influence decision making and therefore both be considered [
15,
16]. Choosing the right behavioral theory is important, but methodological issues must also be addressed. Addressing these barriers could involve measuring decisions on a regular basis over a long period of time to better understand the antecedent factors that lead to a decision, as suggested by Hujala et al. [
14]. This allows a glimpse into the decision-making cycle including length of time between decisions, and what preliminary actions are taken before a decision is made.
In addition to improving current FFO behavior measurement methods, EMAs may provide insight into FFO engagement with their woods. The terms ‘behavior’, ‘action/activity’, ‘decision’, and ‘engagement’ are often used interchangeably in the FFO literature. While behavior and action/activity are nearly synonyms, decisions are more appropriately compared to a behavioral intention. While making a decision can be defined as an outcome of interest, it is more typically the setting of an intention which may or may not manifest in a behavior. Finally, we define natural resource engagement on a spectrum. In forestry, engagement can involve thinking about one’s woods to activities/behaviors like enrolling in a cost-share program or conducting a timber harvest [
17]. Woodland engagement is important for several reasons, including receptiveness to professional advice about best practices and conservation strategies. Engaged woodland owners may not actively manage their woods, but they are potentially open to information and intentionally considering their options, including leaving the forest alone. Policymakers and natural resource professionals can provide better outreach and more efficient programming to engaged woodland owners [
17].
The scholarship has explored traits of more engaged FFOs, but not yet shed light on how to better engage FFOs who might be receptive to a variety of actions, but are not yet undertaking any [
11]. Thus, it is imperative that new methods are explored. Given this need, this study sought to test the EMA method for its potential in understanding FFO decisions and overcoming the methodological issues inherent in traditional self-report retrospective surveys. Woodland owners in the U.S. are older than the general population and tend to live in more rural areas [
8], so we did not restrict our study to smartphone technology, given that there is still a lag in smartphone use among older and rural populations [
18], but we did explore e-mail as a digital advance from traditional mail-based surveys. The overall research goal was to determine whether or not EMAs were reasonable for data collection on FFO behaviors and to measure FFO engagement with an EMA approach. Our specific objectives were to (1) determine if any forest landowners would participate in a real time digital survey process, (2) measure the attrition rate and reaction to real time surveys, and (3) explore levels of FFO activities and engagement using EMAs. Given that the EMA approach has not yet been used in forestry, our hypotheses are based on prior methodological approaches. For example, we expected that FFOs would participate in a digital survey as this approach has been successful, but that we would lose some participants during the process. We expected to find a range of activities undertaken by FFOs, based on the National Woodland Owner Survey [
18], but had no evidence to hypothesize how activities would change on a weekly basis. The National Woodland Owners Survey is administered by the US Forest Service approximately every 5 years and is the most comprehensive source of data on private woodland owners in the United States [
18], focusing on their management actions, ownership objectives, participation in programs, and demographics.
5. Conclusions
This novel method has potential, if expanded, to collect quality data about direct antecedents and the decision-making context for FFOs, shedding much needed insight on what takes place before a forest management decision is made. Over time, the data could show that those landowners who thought about their woods consistently, or engaged in certain activities were more or less likely to engage in other perhaps larger and more consequential activities, such as a timber harvest or selling and subdividing. The actual questions that are asked in an EMA should be adapted to better reflect the antecedents to an important land management decision of interest. As new generations of FFOs take ownership of forest land, it is likely that comfort and familiarity with technology (e.g., smartphones) will increase [
18]. The technique could be deployed via a smartphone application and could include information and links to technical service provider information, making it a dual-purpose research and behavior change tool. An EMA approach could also be used to experiment with outreach and engagement strategies. For example, a FFO who is thinking about their woods versus talking about their woods to others may necessitate different outreach strategies. Additionally, FFOs who are consistently undertaking certain activities compared to those who do activities sporadically or seasonally may need different outreach methods and topics. This novel (in natural resources) method will help shed light on the variability of a single decision-maker through time and the general patterns of decision-making across a broader population, helping natural resource managers better predict the public benefits and services that natural resources provide. Other examples of research topics that could be explored using an EMA include recreation decisions, restoration activities, invasive species control and prevention, defensible space and fire prevention behavior, and a broad range of environmental and sustainability behaviors (e.g., recycling). Finally, the EMA approach could also be applied to other areas of forest science and forest management, such as community-based natural resource management, disturbance ecology, and utility maximization modeling in forest economics. EMAs are particularly useful when (1) rates and frequencies of a behavior or a decision are of interest and (2) when it is important to know what happens for prior to a decision, behavior, or action, particularly if there is a suspected sequence of events or factors that precede this decision or action. It is also useful when complete monitoring is not possible, but temporal factors are critical to measure and understand.