A Control Theory Approach to Understanding the Dynamics of Cognitive Wellbeing

Abstract

(1) Background and Objective: The debate on the stability and variability of subjective wellbeing (SWB) is decades old. However, despite the wealth of literature on this topic, there are relatively few studies that aim to explain the “why” and “how” of the dynamics of SWB. In this context, the objective of this exploratory study is to test the plausibility of a model of the cognitive component of SWB (CWB) inspired by the control theory. In this model, a measure of future life expectations (FLEs) serves as a target in the control loop regulating cognitive wellbeing (CWB), while general self-efficacy (GSE) and affective wellbeing (AWB) are mediators in the direct and feedback loops. (2) Method: To test this model, we collected data from a convenience sample of N-98 Romanian students in Computer Science using well-established questionnaires measuring CWB, GSE, AWB, and FLE. Mediation analyses and path modeling were conducted to evaluate the feedback-based model of the interplay between these variables. (3) Results: The findings confirm a significant relationship between FLE and CWB (β = 0.62, p < 0.001). GSE partially mediates the link between FLE and CWB (β = 0.139, p = 0.02), while AWB mediates the feedback from CWB to FLE (β = 0.297, p < 0.001). The model explains 42% of the variance of CWB. (4) Conclusions: Our study remains exploratory in nature, but preliminary data suggest that a model of SWB based on feedback control is worth attention as it might provide a better understanding of the dynamics of SWB.

1. Introduction

1.1. Conceptual Delimitations

In the past few years, almost five million people enrolled in a course called “The science of well-being” held by Professor Laurie Santos from Yale University on Coursera (Santos 2024), hoping to acquire or improve certain “skills of being happy”. Over the same time period, researchers from all the domains of knowledge published almost 200,000 articles indexed in the Web of Science, with the keyword “wellbeing” in the topic.

The exceptional popularity of this course among the general public reflects the widespread belief that happiness can be taught and learned. However, the dynamics of happiness, to what extent happiness can change at the individual and social levels, and how and why this happens remain a subject of ongoing debate (Cummins 2014; Gibbs 2015; Das et al. 2020; Rusk 2022).

Happiness is an elusive term, often used interchangeably with the related, albeit not synonymous terms, “subjective wellbeing” (SWB), “life satisfaction” (LS), quality of life (QoL), flourishing, and thriving (Rojas and Veenhoven 2013; Oishi et al. 2013; Dodge et al. 2012). It is also complex and multidimensional: a review of the literature by Linton et al. (2016) found 99 self-reporting instruments for measuring wellbeing and no less than 196 dimensions thereof.

To avoid nomenclature confusion, from the many definitions of these concepts found in the literature, for the purposes of this presentation, we will use the following:

Subjective wellbeing (SWB) consists of “good mental states, including all of the various evaluations, positive and negative, that people make of their lives, and the affective reactions of people to their experiences” (OECD 2013).

This definition of SWB encompasses both the affective components described by the term “happiness”, or “affective wellbeing” (AWB), and the cognitive evaluations involved in the concept of “life-satisfaction”, or “cognitive wellbeing” (CWB).

In this paper, we assume that the terms CWB and LS are synonymous.

It results that “SWB can be conceptualized as a momentary state (mood, AWB) as well as a relatively stable trait (LS, CWB)” (Eid and Diener 2004).

In a more comprehensive approach (Kim-Prieto et al. 2005), SWB is described as a process, consisting of the following sequence of stages:

• Perception of life circumstances and events;
• Affective reactions to those events;
• Recall of past experiences;
• Integrative evaluation about one’s life as a whole.

The final stage of this process—the overall evaluation of life—is not limited to past and present experiences: our expectations about the future can greatly affect how we feel now. Among the future related factors proven to influence SWB, we count the following:

• Goals—defined as internal representations of desired outcomes of one’s behavior—can vary in what concerns their content, importance, and the motivation behind pursuing them, but they certainly influence the SWB (Kasser and Ryan 1993; Brdar et al. 2009).
• Hope, defined as the cognitive and motivational capacity to devise and pursue meaningful pathways to desired goals—is a critical determinant of subjective wellbeing because it enhances emotional resilience, fosters positive affect, and ultimately elevates life satisfaction (Pleeging et al. 2021).
• Desires are defined (Crawford Solberg et al. 2002) as personally significant goals or expressed wishes that are chronically accessible and central to life satisfaction. The link between desires and satisfaction lies in “desire discrepancy”, where dissatisfaction occurs when one’s possessions or circumstances do not align with their desires. If desires are unmet, life satisfaction tends to decrease, whereas when desires align closely with reality, satisfaction increases.
• Affective forecasting is defined as the process by which individuals predict their future emotional states in response to anticipated events, including estimations of the valence, intensity, and duration of their feelings. Research by Wilson and Gilbert (2003), Bertoni and Corazzini (2018), and many others indicates that people often misjudge these future affective responses—typically overestimating both the intensity and the persistence of their emotions, and such forecasting errors can significantly impact subjective wellbeing by influencing decision-making and goal-setting. When individuals base major life choices on inaccurate predictions of happiness or distress, the resulting discrepancy between expected and experienced emotions may lead to lower life satisfaction.
• Time perspective. Since all human experiences are processed within a temporal framework, appraisals of one’s own quality of life are also influenced by the individual time perspective. Zimbardo and Boyd (1999, p. 18) defined time perspective as “the often nonconscious process whereby the continual flows of personal and social experiences are assigned to temporal categories, or time frames, that help to give order, coherence, and meaning to those events”. Numerous studies explored the connections between time perspective and subjective wellbeing. A recent meta-analysis of these studies (Diaconu-Gherasim et al. 2023, p. 1) found that “a present hedonist time perspective had positive relations with life satisfaction, happiness, and positive affects. The future time perspective was linked to higher levels of positive indicators of well-being and lower levels of negative indicators”. On the contrary, the present fatalistic time perspective and the deviation from the balanced time perspective are negatively correlated with wellbeing.

In summary, life satisfaction is conceptualized as the integrative outcome of appraising past experiences and present circumstances; nevertheless, it is predominantly determined by future-oriented constructs. Specifically, the anticipation of desired outcomes—as operationalized through goals, hope, and desire discrepancies—alongside individuals’ affective forecasts and temporal perspectives, serves not only to shape the interpretation of past and present events but also to guide adaptive behavior toward attaining a more favorable future. Ultimately, it seems that this forward-looking evaluative process exerts the most profound influence on overall subjective wellbeing.

1.2. The Dual Cognitive–Affective Nature of SWB

Though AWB and CWB are strongly correlated, empirical evidence suggests that they may be considered distinct constructs with different stability with respect to positive and negative life events (Eid and Diener 2004; Kahneman and Deaton 2010; Luhmann et al. 2011, 2012a; Fors Connolly and Gärling 2024). A review and meta-analysis (Luhmann et al. 2012b) on the differential stability of the components of SWB found that persistent changes are more likely to occur in CWB than in AWB.

The interplay between AWB and CWB is still under debate. It is not clear whether the prevalence of positive affect over a certain time period results in higher life satisfaction, or whether people more satisfied with their lives tend to reframe the negative events or circumstances more positively. For example, unemployed persons have significantly lower life satisfaction than employed persons, but their daily AWB is comparable (Knabe et al. 2010).

Also, AWB and CWB appear to have different determinants (Diener et al. 2010; Fors Connolly and Gärling 2024) and different consequences (Wiest et al. 2011; Gan 2020).

1.3. Brief Overview of the Ideas on the Dynamics of SWB

Many of the early studies on SWB tended to overestimate its stability. Brickman (1971) described a process called “hedonic adaptation” similar to sensory adaptation, which makes people quickly adapt to positive and negative life events (e.g., winning the lottery, having a serious accident, etc.) and return to a state of hedonic neutrality. As a result, the pursuit of happiness is like a run on a “hedonic treadmill”.

The baseline level of SWB, called the “set point” (Headey and Wearing 1992), corresponding to hedonic neutrality is believed to be genetically determined. Lykken and Tellegen (1996, p. 186) noticed that separated twins who grew up in different families had similar levels of SWB, and argued that “neither socioeconomic status, educational attainments, family income, marital stalls, nor an indicant of religious commitment could account for more than about 3% of the variance in WB. From 44% to 52% of the variance in WB, however, is associated with genetic variation”.

Other studies linked SWB with personality traits such as extraversion and neuroticism, and a meta-analysis of these studies (Steel et al. 2008, p. 138) concluded that “total SWB variance accounted for by personality can reach as high as 39% or 63% disattenuated”.

A more elaborated attempt to explain the remarkable long-term stability of SWB is the theory of SWB homeostasis (Cummins 1995; Cummins 2010; Cummins et al. 2012), which starts from the analogy between the physiological mechanisms that control certain body states (such as blood pressure, body temperature, blood glucose, oxygen concentration, etc.) and the psychological management of SWB. According to this theory, the stability of SWB is automatically maintained by means of certain external and internal “buffers”, as shown in Figure 1.

Cummins (2010) argues that the set point of SWB is genetically determined and states that, just like in other homeostatically controlled processes, if SWB drops too far below the baseline, it can overwhelm the system’s ability to maintain balance, leading to the defeat of homeostasis.

However, analogy does not mean identity. The idea of a universal, genetically determined set point of SWB is questionable. While all people have the same body temperature, there are obvious inter-individual differences in what concerns the SWB. It appears that at least some components of the SWB set points are individual and may change in time.

More recent studies (Lucas et al. 2004; Lucas 2007) show that “although inborn factors certainly matter and some adaptation does occur, events such as divorce, death of a spouse, unemployment, and disability are associated with lasting changes in SWB. These studies also show that there are considerable individual differences in the extent to which people adapt. Thus, happiness levels do change, and adaptation is not inevitable” (Lucas 2007, p. 75). And longitudinal studies confirm this idea. For example, the findings in Kettlewell et al. (2020) provide insight into the temporal dynamics of AWB and CWB (see Figure 2 and Figure 3 for an illustration of the impact of positive and negative life events on AWB and CWB).

Though widely accepted for decades, the adaptation theories of SWB, including the homeostasis theory, have been criticized for several important reasons. A detailed critique of these theories is available in (Luhmann and Intelisano 2018). If they were true, then any “individual and societal efforts to increase happiness are doomed to failure” (Diener et al. 2009), and “trying to be happier is as futile as trying to be taller” (Lykken and Tellegen 1996, p. 189). This conclusion is obviously counterintuitive, and a growing number of studies, mainly coming from positive psychology, describe interventions demonstrating that lasting improvements of SWB are possible. See (Bolier et al. 2013; Carr et al. 2021) for a systematic review and meta-analysis of these studies.

Some of the most important drawbacks of the adaptation theories of SWB according to (Bolier et al. 2013; Headey 2007, 2010; Diener et al. 2015) are as follows:

• The actual control mechanisms involved in the stability of SWB are not sufficiently explained. The concept of “SWB buffers” used in (Cummins 2010) is rather elusive.
• These theories do not explain the individual differences in the adaptive response time of the control loop of SWB.
• The stability of SWB may be the result of multiple control loops, with multiple set points.
• Set points do not necessarily correspond to hedonic neutrality, but tend to be, on average, positive.
• Finally, these theories do not explain the lasting positive effects on SWB obtained, for example, through positive psychology interventions (PPIs) (Bolier et al. 2013).

Facing these critics, Robert Cummins, the most prominent supporter of the homeostasis theory of SWB, acknowledged that only the affective components of SWB, which he called “homeostatically protected moods (HPMoods)”, are subject to homeostatic control (Cummins 2010, p. 10) and states that “HPMood is not only the dominant affective constituent of SWB, […], but also the basic steady-state, set-point that homeostasis seeks to defend”.

Besides the adaptation theories of SWB, the affective forecasting theory provides valuable insights into the dynamics of subjective wellbeing. Defining affective forecasting as the mechanism by which people attempt to foresee their future feelings, Bertoni and Corazzini (2018) note that the effect of forecasting errors on SWB is asymmetric: unmet expectations (i.e., negative forecasting errors) are significantly associated with lower life satisfaction, whereas exceeding expectations (i.e., positive forecasting errors) does not confer a comparable benefit. This finding links the dynamics of SWB with loss aversion principles (Tversky and Kahneman 2000) and helps explain the asymmetrical response of SWB to certain positive and negative life events (Luhmann et al. 2012b) as well as the inter-individual differences in SWB. However, the affective forecasting theory, like the adaptation theories, fails to address the role of individual agency—namely, the conscious behavioral changes that might minimize forecasting errors. Thus, the explanatory value of affective forecasting theory lies in illuminating the sources of variability in subjective wellbeing rather than in explaining its stability.

The conclusion is that “after assuming for decades that SWB is highly stable, the field has recently made a significant turn towards recognizing that lasting changes in SWB are possible” (Luhmann and Intelisano 2018, p. 18).

However, the idea of the existence of certain genetic factors that contribute to the stability of SWB is not totally rejected. Diener et al. (2018, p. 2) estimate that “on average, about 30 to 40% of the variance in individual differences in SWB is attributable to genetic effects”. It results that the remaining 60–70% of the variance is due to other factors. “Thus, there are likely many controllable factors that can increase or lower SWB”.

2. Towards a Model of CWB Based on Feedback Control

2.1. Control Theory as a Framework for Understanding the Dynamics of SWB

Natural systems are inherently dynamic, subject to random environmental perturbations that, in accordance with the second law of thermodynamics, drive these systems toward increasing disorder. The persistence of stability in certain phenomena suggests the operation of self-regulatory mechanisms that counteract these destabilizing forces. In particular, the notable stability of SWB, as discussed in the previous section, may be more readily explained from the perspective of self-regulation.

The theoretical framework for investigating self-regulation is control theory—a branch of engineering dedicated to understanding the behavior of dynamic systems and devising methods to influence that behavior to achieve desired outcomes. In such systems, when an output variable is required to follow a specific reference (or set point) over time, the system’s input is modulated through negative or positive feedback loops to secure and maintain the desired output despite external perturbations (see Figure 4).

The most intuitive example is the thermostat—a feedback-controlled system that continuously measures the temperature, compares it to a reference value, and adjusts the behavior of the system to minimize the difference.

The idea of using the control theory as a framework for understanding certain psychological processes, emphasizing their dynamic and adaptive nature, is not new. One of the earliest and most influential proponents was William T. Powers, who introduced Perceptual Control Theory (PCT) in the 1970s (Powers 1973). Powers proposed that behavior is the control of perception, suggesting that individuals act to keep their perceptions of the world within acceptable bounds.

Control theory was further integrated into psychological research through the work of Carver and Scheier (1982), who applied it to self-regulation and goal-directed behavior. Their model emphasized the role of feedback in maintaining goal-directed behavior and has been influential in research on motivation, emotions, and behavior.

Later, Carver and Scheier (1982) explained positive and negative affect as a result of a feedback-controlled process. Positive affect occurs when the rate of progress toward achieving goals exceeds the desired standard or expectation, signaling success or advancement toward a goal. Negative affect arises when the rate of progress falls below the expected standard, indicating slow or insufficient progress.

Carey et al. (2009) used PCT as a framework to conceptualize psychological distress as resulting from conflicts within control systems, and introduced the Method of Levels (MOL)—a psychotherapeutic approach derived from PCT—to help patients resolve these conflicts.

Finally, Ferguson et al. (2018) used PCT to frame and analyze how feedback interventions affect prescribing behaviors among junior doctors by examining how discrepancies between prescribers’ reference values (i.e., their standards for appropriate prescribing) and their actual performance trigger behavioral adjustments.

See Mansell and Marken (2015) for a review of the literature on the use of control theory in psychology.

Though not always explicitly stated, some of the existing attempts to explain the dynamics of SWB are inspired by the control theory. The term “set point”, defined as the target value or range of values that a system aims to maintain through regulatory mechanisms, is borrowed from control theory. Also, the term “forecasting error” used in the affective forecasting theory is very similar to the concept of “control error” from control theory.

The biological process of homeostasis is a typical example of a feedback-controlled process. For example, the homeostatic process of maintaining body temperature (the controlled variable) within a specified narrow interval involves sensors (thermo-receptors—nerve endings located mainly in the skin that detect temperature variations), a control center (hypothalamus), and a number of effectors (sweat glands activated to produce sweat, which cools the body through evaporation, blood vessels that dilate or constrict to adjust the blood flow in the skin, muscles that may generate heat through shivering, and hormones like thyroxin, which increases the metabolic rate).

In summary, the idea that certain psychological processes, including SWB, can be better explained in the framework of the control theory seems compelling. In this view, small variations in SWB are due to disturbances (like life events), and long-term modifications correspond to changes in the set point.

2.2. Building Hypotheses

Building upon the similarity between the evolution of the output of a feedback-controlled system (shown in Figure 4b) and the evolution of SWB measured in longitudinal studies (Figure 2 and Figure 3), the general objective of the present study is to explore the plausibility of a model of SWB as a feedback-controlled system.

To this purpose, we started with the following theoretical assumptions derived from the existing literature briefly reviewed in the previous section:

• SWB is not a state but a process of continuous adjustments (Kim-Prieto et al. 2005).
• The process of adjustment of the SWB towards the target value is not entirely unconscious. People can and do set conscious goals, actively make decisions, and adjust behavior in order to reach these goals (Hitlin and Kirkpatrick Johnson 2015).
• Though AWB and CWB are strongly correlated, they are distinct constructs (Luhmann et al. 2012a) and their distinct levels of stability are due to the existence of distinct feedback control loops. While the set point of the control loop for AWB seems to be genetically determined (Cummins 2010), we hypothesize that the control loop for CWB follows a different target value, which is individual and may change with time.

Based on these ideas, we hypothesize that the control loop for CWB has the structure shown in Figure 5.

In the next step, we formulated hypotheses about the psychological variables that may act as control loop target, or mediate the link between the target and life satisfaction.

Regarding the control loop target, we noticed that hope, goals, and a future-oriented time perspective are encompassed by the broader concept of “future life expectations” (FLEs), which summarize one’s beliefs and assumptions about one’s future life.

FLEs are fairly stable in time, albeit certain variations due to external life events are still possible. It appears that a measure of FLEs may be a good candidate for the target of the CWB control loop.

Hence, the primary research hypothesis of the present study can be formulated as follows:

H1. 

There is a significant correlation between FLEs and CWB, suggesting a bidirectional interplay between them.

As shown in Figure 5, according to the control theory, there are two distinct pathways between the target of the control loop (FLEs) and the controlled variable (CWB), suggesting that each pathway may involve a separate mediating mechanism.

A possible mediator between FLEs and CWB is self-efficacy, defined as people’s confidence in their ability to control their functioning and the events that impact their lives (Bandura 1977). This idea is supported by previous research linking GSE with SWB and LS (Cattelino et al. 2023; Hussein Alkhatib 2020; Yilmaz 2018; Savi Çakar 2012). The connection of GSE with positive life expectations has been explored by Kim (2014), while other studies identified strong correlations of GSE with optimism (Karademas et al. 2007; Yu and Luo 2018), hope, and positive expectations (Gallagher et al. 2020). GSE is also correlated with the internal and external “buffers” mentioned by Cummins (2010) as factors of the stability of SWB, namely self-esteem (Chen et al. 2004), social support (Karademas 2006), and a sense of control over life events (Amaral et al. 2024; Tavousi et al. 2009).

Thus, our second research hypothesis can be formulated as follows:

H2. 

Self-efficacy partly mediates the effect of FLEs on CWB.

In the general model of a feedback-controlled system shown in Figure 4, the feedback loop always comprises a measurement subsystem that provides information about the current status of the output. In other words, the information considered in the feedback loop describes the output but differs in terms of the time frame used for evaluation: while the output is assessed from the perspective of the global behavior of the system, the information obtained through measurement describes the momentary status thereof. On the other hand, AWB can be seen as a momentary affective reflection of the general life satisfaction, taking into account the current person’s interaction with the environment (Carver and Scheier 1982; Tov 2018).

Therefore, we hypothesize the following:

H3: 

AWB partly mediates the effect of CWB on FLEs.

The entire model of the interdependencies between FLEs and life satisfaction built according to the above hypotheses is shown in Figure 6.

3. Method

To test these hypotheses, we conducted an exploratory study with a small sample of Romanian undergraduate students in Computer Science, mainly aiming to identify variables and potential relationships between them rather than attempting to draw definitive conclusions.

3.1. Participants and Procedure

After being properly informed about the objectives and procedures of the study, a sample of N = 98 undergraduate students at the Department of Computers and Information Technology of the University “Dunarea de Jos” of Galati, Romania, aged [20–39], M (SD) = 21.57 (3.03), 65 males (66%) and 33 females (34%), voluntarily enrolled to participate in the study. Data were collected online between February and April 2024.

3.2. Instruments

All participants completed a Google Forms questionnaire with distinct sections for each observed variable. Besides the basic demographic information, an additional question was introduced to assess the participants’ English proficiency. All the participants’ reported English proficiency was above average. This was required because, in order to avoid any possible biases introduced through translation, all the scales were presented in their original form in English. The responses were collected by means of a 5-point Likert scale ranging from 0 (not at all/strongly disagree) to 4 (very much/completely agree).

To measure the main variables, we chose to use well-established scales with demonstrated validity.

Life satisfaction was measured with the Personal Wellbeing Index (PWI) (International Wellbeing Group 2013), which comprises seven questions of the type “how satisfied are you with your health”, “your standard of living”, etc.

AWB was measured with the 5-item World Health Organization Wellbeing Index (WHO-5) (Staehr 1998), which captures the affective aspects of SWB in the recent past and consists of statements of the type “over the past two weeks … I have felt cheerful and in good spirits, etc.”.

Self-efficacy was measured with the Generalized Self-Efficacy Scale—GSE (Chen et al. 2001)—consisting of 8 items of the type “I believe I can succeed at most any endeavor to which I set my mind”.

Future life expectations (FLEs) were assessed with a simplified version of the Aspirations Index (Kasser and Ryan 2001), consisting of seven items of the type “Considering a 10-year time horizon, estimate how likely you believe the following objectives are to be achieved: I will have a happy family” (see Appendix A).

The internal consistency values of the scales obtained with our sample are shown in

Table 1. Cronbach alpha coefficients for the scales used in this study (N = 98).

Variable	Scale	Number of Items	Cronbach Alpha
Life satisfaction	PWI	7	0.778
AWB	WHO-5	5	0.879
General self-efficacy	GSE	8	0.900
Future life expectations	FLE	7	0.827

.

3.3. Data Analysis

Descriptive statistics analysis was conducted using SPSS 20. For the mediation analyses, we used model 4 of the Process Macro v.4.2 (Hayes 2017). The path analysis for the entire closed-loop model was conducted using Amos Graphics (Arbuckle 2011).

Descriptive statistics and the correlation matrix for the main variables are shown in

Table 2. Descriptive statistics and correlation coefficients for the main variables (N = 98).

	Mean	SD	PWI	WHO-5	FLE
PWI	19.22	4.587	1
WHO-5	10.79	4.616	0.718 ***	1
FLE	20.57	4.922	0.620 ***	0.541 ***	1
GSE	23.22	5.160	0.519 ***	0.377 ***	0.585 ***

*** Correlation is significant at p < 0.001 (two-tailed).

.

4. Results

To test hypotheses H1 and H2, we conducted a mediation analysis with FLEs as the independent variable (X), PWI as the dependent variable (Y), and GSE as the mediator (M). The results are synthetically presented in

Table 3. Results of the mediation analysis for the path FLEs → GSE → PWI.

Effect	B	SE (B)	β	t	p
a: FLEs → GSE	0.6138	0.0867	0.5855	7.0756	0.0000
b: GSE → PWI	0.2110	0.0855	0.2374	2.4669	0.0154
c (total): FLEs → PWI	0.5780	0.0746	0.6203	7.7478	0.0000
c’ (direct): FLEs → PWI	0.4485	0.0897	0.4813	5.0011	0.0000
ab (indirect): FLEs → GSE → PWI	0.1295	0.0560	0.1390	2.3125	0.0210

.

To test hypothesis H3, we performed mediation analysis with WHO-5 as a possible mediator between PWI (X variable) and FLEs (Y variable). The results are presented in

Table 4. Results of the mediation analysis for the path PWI → WHO-5 → FLEs.

Effect	B	SE(B)	β	t	p
a: PWI → WHO-5	0.4213	0.0662	0.5449	6.3677	0.0000
b: WHO-5 → FLEs	0.6578	0.0940	0.5457	7.0000	0.0000
c (total): PWI → FLEs	0.5781	0.0746	0.6203	7.7478	0.0000
c’ (direct): PWI → FLEs	0.3009	0.0727	0.3229	4.1419	0.0001
Ab (indirect): PWI → WHO-5 → FLEs	0.2770	0.0590	0.2974	4.6949	0.0000

.

Figure 7 synthesizes the standardized regression coefficients for the two models.

Finally, we analyzed the closed-loop model described by the path diagram depicted in Figure 8 using Amos Graphics.

The model fitness was evaluated as proposed by Hu and Bentler (1995) and Byrne (2013) by comparing the values of the Chi-square/degrees of freedom ratio (χ²/df), Standardized Root Mean Square Residual (SRMR), Root Mean Square Error of Approximation (RMSEA), Goodness of Fit (GFI), Tucker–Lewis Index (TLI), and Comparative Fit Index (CFI) with the thresholds listed in

Table 5. Threshold values of the main model fit indices.

Index	Description	Threshold
χ2/df	Chi-square divided by degrees of freedom	≤3
GFI	Goodness of Fit Index	≥0.90
TLI	Tucker–Lewis Index	≥0.90
CFI	Comparative Fit Index	≥0.90
RMSEA	Root Mean Square Error of Approximation	≤0.06
PCLOSE	p-value of closeness of fit	≥0.05

.

The results of the path analysis for the model, shown in Figure 5, indicate a very good fit, as shown in

Table 6. Fit indices for the model shown in Figure 8.

Index	Description	Value
χ2/df	Chi-square divided by degrees of freedom	0.378
GFI	Goodness of Fit Index	0.998
TLI	Tucker–Lewis Index	0.987
CFI	Comparative Fit Index	1.000
RMSEA	Root Mean Square Error of Approximation	0.000
PCLOSE	p-value of closeness of fit	0.585

.

The values of the regression weights, presented in

Table 7. Regression weights.

			Estimate	S.E.	C.R.	p
PWI	<---	FLE	0.391	0.119	3.298	***
WHO	<---	PWI	0.469	0.057	8.226	***
FLE	<---	WHO	0.343	0.161	2.134	0.033
GSE	<---	FLE	0.582	0.089	6.552	***
PWI	<---	GSE	0.253	0.098	2.586	0.010

*** p < 0.001.

, all have a critical ratio (C.R.) above the threshold of 1.96, indicating that the magnitudes of the regression weights for all paths in the model are significant with respect to the standard errors (Hayes 2017).

5. Discussion

The correlation between FLEs and PWI is high: Pearson r = 0.62, p < 0.001. The direct effect of FLEs on PWI is significant (β = 0.48, p = 0.000) and stronger than the direct effect of the reverse path PWI → FLEs (β = 0.33, p = 0.001). The total effect is significant in both directions (β = 0.62, p = 0.000), indicating that hypothesis H1 is verified.

The mediating effect of GSE between FLEs and PWI is weak but statistically significant (β = 0.139, p = 0.02); thus, hypothesis H2 is also confirmed.

The mediating effect of AWB between life satisfaction and FLEs is stronger (β = 0.297, p < 0.001) and important from a theoretical perspective, because it supports the idea of the existence of a feedback link between CWB and FLEs. Hypothesis H3 is confirmed.

Though very simple, this model explains 42% of the variance of life satisfaction (see Figure 8), a ratio that seems important, considering the fact that the model does not include other factors proven to impact life satisfaction, such as cultural and regional factors (Lai et al. 2013; Margolis and Myrskylä 2013), balance between work and family life (Lindfors et al. 2007), perceived control (Wardle et al. 2004), or religion (Tay et al. 2014).

The proposed model is consistent with previous studies. For example, Park and Suh (2023, pp. 1–2) found a significant positive correlation (r = 0.669, p < 0.001) between life satisfaction and expectations for future life and suggest that the link is bidirectional: “expectations for a satisfied life in the future should be included when measuring subjective well-being […] there is also evidence that if individuals are satisfied with their present lives, they may be more likely to have higher expectations for their future lives”. Shan et al. (2022) found a significant correlation between life satisfaction and psychological capital—a complex construct that includes self-efficacy, hope, optimism, and resilience. Many other studies explore the link between GSE and LS. For example, in Savi Çakar (2012), it is argued that GSE explains as much as 48% of the variance of LS, and in Azizli et al. (2015), GSE is found to correlate with LS and with the scores on the Continuous Planning Scale (CPS). The connection of GSE with positive life expectations is explored by Kim (2014).

The link between AWB and LS has been extensively studied (see, for example, Kettlewell et al. 2020; Luhmann et al. 2012b; Fors Connolly and Gärling 2024) as they are both definitory dimensions of SWB.

Though it makes intuitive sense that the predominance of positive or negative affect may influence behavior and cognitive processes, there are very few studies explicitly aiming to examine the interplay between AWB and FLEs. Most of the existing studies (e.g., Chopik et al. 2020; Yoshimura and Hashimoto 2020) are concerned with the related construct of optimism.

Thus, from a theoretical perspective, the model proposed in our study appears to be plausible.

This model allows for a better understanding of the long-term stability as well as the short-term and inter-individual variations in CWB. It also explains the efficiency of certain positive psychology interventions, such as the best possible self-exercises (Carrillo et al. 2019; Orkin et al. 2023), which consist of asking the subjects to envision their ideal future self in the absence of adverse circumstances, and write about it in detail. This creates a clear inner picture of the personal aspirations across different domains like career, relationships, financial security, and social status, fostering higher life expectations, which further lead to lasting improvements in SWB.

Of course, this study remains exploratory in nature. The most important and obvious limitation is that our sample is too small and not representative of the respective age group. At least in Romania, students in Computer Science are almost certain that they will quickly find a well-paid job right after graduation. Many of the participants in our study were already employed at the time of completing the questionnaire. Hence, their future life expectations tend to be higher than the average, at least concerning their job satisfaction and financial situation. Moreover, the fact that the questionnaires used were only available in English, while the participants were non-native English speakers, may introduce another potential response bias; however, this bias is considerably smaller than the bias associated with translation.

Future research should replicate this study with more diverse samples and in different cultural contexts. Other age groups should also be considered in order to elucidate whether the decline of SWB in late adulthood is linked with lower FLEs (Buecker et al. 2023).

6. Conclusions

This exploratory study found that a model of CWB inspired by the control theory in engineering is plausible. In this model, FLEs define the target value of the control loop and GSE and AWB are mediators of the direct and feedback link between FLEs and CWB. This new model suggests that life satisfaction can be redefined as “the result of the evaluation of the extent to which one’s current life is congruent with one’s life expectations”.

From a theoretical perspective, the proposed model supports the idea that CWB and AWB are distinct constructs—each regulated by its own feedback loop—and provides a framework for better understanding both the relative stability and variability of SWB.

Practical implications include the possible design of interventions that either directly aim to foster life expectations or reduce the influence of negative affect on FLEs. Knowing that positive life expectations do predict better health (Lench 2011), this might be of interest for policy makers in public health and education.