Differentiating hope from optimism by examining self-reported appraisals and linguistic content

ABSTRACT

Even though hope and optimism are both positive states about desired future events, we hypothesized that their appraisals differ. Specifically, we hypothesized that hope would be associated with greater appraisals of uncertainty, importance, effortful action, morality, unpleasantness, and fear than optimism. Based on action identification theory, we also hypothesized that hope would encourage using more concrete language than optimism. In three experiments, respondents wrote about possible future events that instilled feelings of either hope or optimism. We assessed appraisals via respondents’ self-reports and by coding events for appraisal-relevant language. An internal meta-analysis of three experiments revealed that, compared to optimism, hope involved more uncertainty, importance, effortful action (self-reports only), unpleasantness, fear, and concrete language, but not more morality. These data suggest that even though hope arises when the distal future seems more uncertain and unpleasant, hope might help people obtain their goals by signaling importance, effort, and promoting concrete thinking.

KEYWORDS:

• Hope
• optimism
• emotion
• appraisal
• concrete language

Acknowledgments

We’d like to thank Margaret Aichele, Jennifer Heyman, Karan Khindri, Taylor Kline, Amanda Kuroda, Samantha Mayhew, Nancy Nwogu, and Shanyé Phillips for their help with data collection.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental Materials

Supplemental data for this article can be accessed here

Notes

1. Some researchers might question whether optimism is an affective state, because it sometimes is defined as an expectation. This expectation, however, typically involves an affective reaction (Bruininks & Malle, 2005). Indeed, Clore et al. (1987) classified optimism as an affective condition, because one can both feel optimistic and be optimistic. Shaver et al. (1987) even defined optimism as an emotion, signaling that ‘…one expects a joy-producing outcome in the future’ (p. 1069). Thus, optimism can be an affective state.

2. Base rate data were calculated by averaging across six feelings (hope, optimism, want, desire, wish, and joy) the number of participants (out of 52) that mentioned taking action.

3. Experiment 1 also measured appraisals of control and appropriateness, but these measures were problematic (see supplemental materials).

4. In Experiment 3, we added ‘To what extent do you feel that it is morally right to be hopeful/optimistic that this event will happen?’ which resulted in a poor alpha, α = .43. We dropped this item resulting in a more reliable measure that directly replicates Experiment 1.

5. We also tested whether recalling a hopeful or optimistic event might produce incidental effects on downstream measures of abstract thought; specifically, using abstract descriptions to describe others’ actions, estimating how quickly time passes, and categorization of unusual exemplars. Because these variables focus on the downstream consequences of hope and optimism and are not relevant to appraisal processes, we discuss them in the supplemental materials.

6. In Experiment 2, we examined whether story order, if the story was recalled first or last, altered the results. Order did alter the data, but these order effects did not depend on whether the event involved hope or optimism. Thus, we collapsed across order to simplify the data and describe the order effects in the supplemental material.

7. Because temporal distance was an ordinal scale, we also conducted analyses treating it as such to confirm that the data produced similar values as the t-tests. As predicted, hopeful events took place further in the future than optimistic events. An ordinal logistic regression revealed that the odds of choosing a longer time period were higher in the hope condition versus the optimism condition, Experiment 1: Exp(B) = 1.52, 95% CI [1.09 to 2.12], a significant effect, Wald χ2(1) = 5.97, p = .015, Experiment 3: Exp(B) = .71, 95% CI [.51 to 1.00], a significant effect, Wald χ2(1) = 3.84, p = .05; Experiment 2: Wilcoxon Signed-Rank Test indicated hopeful events (Median = 7) would take longer to resolve than optimistic events (Median = 6), Z = -3.56, p < .01. Thus, even when the ordinal nature of this measure is taken into account, hopeful events are estimated to take place further in the future than optimistic events.

8. The data were skewed, with many people reporting high amounts of both these emotions. Given this skew, we also conducted nonparametric analyses on these data, which revealed the same pattern of results as the t-tests.