NOTE-1316: The impact of place and legacy framing on climate action: A lifespan approach

Contribution

A replication and extension of Zaval et al.’s (2015) legacy induction in an age-balanced sample, beside a new place-based induction. It asks whether older adults, said to care more about legacy and place, respond more. They do not: both inductions work about equally at every age.

Key insight

Framings that tie climate action to what outlasts the writer work, but not by separate routes. The authors expected the legacy task to act through legacy motivation and the place task through love of nature. Each raised both, and the authors grant both tasks may have moved “an underlying third element” (p. 13).

Assumptions

• Legacy motivation is what the Legacy Motivation Scale measures. It is described as “how much participants care about their legacy” (p. 4). Its eight items (S1 File) are:

  • wanting to be remembered: “I care about what future generations think of me”; “It is important to me to leave a positive legacy”;
  • self-regard and efficacy: “I have important skills I can pass along to others”; “I am good at many things”; “I am well liked by my friends”; “Others would say that I have made unique contributions to my community or society”;
  • caring about future people: “I feel a sense of responsibility to future generations”; “I feel a connection to future generations”.

  Only the total is analysed.

• A hypothetical donation proxies behaviour. Participants split a $10 lottery bonus, which one participant would win, between themselves and Trees for the Future, described as helping families “for generations to come” (S1 File).

• Baron and Kenny mediation with one measured mediator at a time, checked by bootstrap (pp. 8–10).

Key results

• Sample (pp. 3–4). 1,005 recruited on TurkPrime in December 2016, five quota groups of 200 by age; 17 excluded on attention checks, leaving 988. Analysed as young (18–35, n = 327), middle (36–52, n = 324) and older (53–87, n = 337). 68% women; 45% called themselves environmentalists. Conditions: Place 299, Legacy 295, Control 394.
• The inductions (p. 5; S2 File).
  • Legacy: “think about what you want future generations to remember you for when you’re gone”, and about “ways in which you will have a positive impact on the environment or on other people”, with examples of saving water and energy, cutting emissions and supporting environmental laws.
  • Place: a place one has an emotional connection to, the impact of climate change on it, and “What would you like that place to be like for future generations after you’re gone?”
• Manipulation checks (p. 6).
  • Legacy motivation: induction F(2, 967) = 20.62, η² = .041. Both tasks beat control (Place p = .001, Legacy p < .001); Legacy beat Place (p = .002). Young adults lower than the rest (η² = .008). No interaction (p = .793).
  • Biophilia: induction F(2, 979) = 8.12, η² = .016. Both beat control (Place p = .002, Legacy p < .001) and did not differ. Rises with age (η² = .041).
• Intentions (p. 6; Table 1, p. 7). Induction F(2, 967) = 15.04, η² = .030; both beat control, not each other. Older adults higher (η² = .012). No interaction (p = .299). Cell means: control 3.27–3.35, Legacy 3.64–3.89, Place 3.35–3.85.
• Donations (p. 7; Table 2, p. 8). Induction F(2, 979) = 4.30, p = .014, η² = .009; Place p = .006, Legacy p = .027 against control. Young adults gave less (η² = .010). No interaction (p = .478). Means from $2.90(young,control)to$4.54 (older, Legacy).
• Mediation, one mediator (pp. 8–10).
  • Legacy → intentions: c = .203, a = .248, b = .355, c′ = .118 (partial); bootstrap indirect 0.180, 95% CI [0.120, 0.245].
  • Legacy → donations: c = .083 (p = .029), b = .139, c′ = .054 (p = .171, “fully mediated”); indirect 0.069 [0.028, 0.118].
  • Place → donations and intentions through biophilia: both partial.
• Mediation, both mediators (S3 File).
  • Legacy → intentions: through biophilia 0.060 [0.016, 0.107], through legacy motivation 0.107 [0.065, 0.155]; direct c′ = 0.135, p = .039.
  • Legacy → donations: through biophilia 0.138 [0.038, 0.245], through legacy motivation 0.097 [0.014, 0.198]; direct c′ = 0.035, p = .869. Total c = 0.270, not significant.
  • Place: no indirect effects through either.
• Correlations (S3 File, Table S1). Legacy motivation with biophilia .396, with intentions .398, with donation .175.
• Nulls (p. 12; S4 File). Neither task changed death-thought accessibility (p = .725), hazard belief or worry. Future self-continuity on overlapping circles, self now and in five years, did not move with induction (p = .201); an age × induction interaction (p = .050) put older adults in the Legacy condition highest.

Claims

id	claim	strength	support
C1	Writing about one’s legacy, or about a loved place under climate change, raises pro-environmental intentions and hypothetical donations against a neutral task	moderate: N = 988, randomized; small effects; self-report and hypothetical giving	pp. 6–8, Tables 1–2
C2	These effects do not differ by age	moderate: no interaction in either outcome; a null with about 100 per cell	pp. 6–7
C3	Legacy motivation mediates the legacy task’s effect	weak: one-time cross-sectional mediator measured after the task; the mediator also moved under the place task; love of nature mediates in parallel	pp. 8–9; S3 File
C4	Biophilia mediates the place task’s effect	weak: significant alone, not in the model with both mediators	pp. 9–10; S3 File
C5	Legacy framing does not work through thoughts of death	weak: a null on one word-completion measure	S4 File

Concepts

• legacy motivation: as scaled here, a mix of wanting to be remembered well, believing one has something to pass on, and feeling responsible to and connected with future generations.
• place attachment: emotional and cognitive bonds with a particular place (p. 2); not measured, only induced.
• biophilia: affiliation with and care for nature, here five items including “Protecting the wellbeing of nature for its own sake is important to me”.
• future self-continuity: felt overlap of self now with self in five years, on Ersner-Hershfield et al.’s circles.

Connections

• Zaval, Markowitz and Weber 2015 (LIT-1724), unread. The source of the induction, the Legacy Motivation Scale, the intentions scale and the donation measure. See section 3.
• Wade-Benzoni (NOTE-1273) never measures legacy concern and generosity in one study. This one does, but with a mixed scale.
• Hurlstone et al. (NOTE-1279) bundle identification with generations, a moral appeal and one’s death. The legacy task here bundles being remembered, one’s own impact and examples of the very behaviours then measured.
• Hershfield (NOTE-1259). The future-self circles come from his group. Legacy framing did not move them in general.
• Many Labs 4 (NOTE-1322). Legacy framing did not raise death thoughts, so it gives terror management no foothold here.

Bearing on the record

• The legacy question (QUESTION-301). Its “evidence from the legacy experiments” asks whether legacy priming acts through the wish to be remembered. This study cannot say. These are inferences from the printed items and results.
  • The mediator is a total over eight items. Two are about being remembered, four about self-regard, two about responsibility and connection to future generations. “Mediated by legacy motivation” is mediated by that mix.
  • The legacy task also raised love of nature, and in the two-mediator model love of nature carried more of the donation effect (0.138 vs 0.097). Concern for nature “for its own sake” runs in parallel with any self-regarding route.
  • The place task, which asks about a place “for future generations after you’re gone” and never mentions being remembered, also raised legacy motivation against control (p = .001), though not in the two-mediator model (a = .032). The score rises without the wish to be remembered being prompted.
  • No condition varies whether anyone will know what one did, which is the question’s test between the thesis’s or Pitcher’s reading and Scheffler’s.
• The triviality objection (CLAIM-492). Legacy motive, as measured, is not a criterion that can come apart from caring: caring about future generations is inside the scale. The one measure of a self extended in time, future self-continuity, did not move with legacy framing except among older adults, and was not tested as a mediator.
• The owner’s thesis (CLAIM-491). Neutral. The study shows that thinking about what outlasts one raises concern for the future; it does not say whose good is served. No dated note on the claim; a dated note on QUESTION-301 is proposed (section 4).
• No THEORY filed.
• No ML instruction.

Limitations

• Self-report and hypothetical giving. Donations were from a bonus one participant might win; nobody’s behaviour was observed.
• Demand. The legacy prompt lists the behaviours the intentions scale then asks about.
• Mediator measured after the task, with the outcomes. Order of the battery is not reported; causal mediation is not established.
• No place-attachment check, as the authors grant (p. 13).
• Timing. December 2016, just after the US election; the authors note it (p. 13). US participants only.
• Unequal cells (Control 394 against about 300) and shifting degrees of freedom (967 vs 979) are not explained.
• Inconsistencies.
  • The intentions scale is reported as “Range = 0–6, M = 2.73” (p. 4), but S1 scores it 1–6 and every cell mean in Table 1 lies between 3.27 and 3.89. The overall mean cannot sit below every cell; it may be on a 0–5 coding (an inference). The Legacy and biophilia ranges (“0–6”, “0–7”) likewise differ from S1’s 1–6 and 1–7.
  • Each bootstrap indirect effect is about twice the product of the stepwise coefficients beside it (intentions: .248 × .355 ≈ .088 against 0.180). The units are not stated.
  • The Legacy effect on donations is significant in the main text (p = .029) and not in S3 (c = 0.270, ns).
  • S3’s last paragraph repeats the donation model’s direct effect (c′ = 0.319, p = .127) for the intentions model, whose figure gives 0.050.
  • “Bartlett’s test indicated sphericity assumptions were adequately met” (p. 6) in a between-subjects design.
• Typos. Reference 26 gives Zaval et al.’s pages as “26(2):3–19” for 231–236 (p. 16); “inexorably linked” for “inextricably” (p. 2); “Zaval and colleague’s” (p. 12); “a $10 dollar lottery” (p. 4); “The potency of our findings imply” (p. 12); “hp2” for ηp² (S4 File).

Open questions

• Does legacy framing work through being remembered? A study that scores the two remembering items apart from the two caring items, and adds an anonymous condition in which one’s actions will never be known, would answer QUESTION-301’s test.
• Does the legacy task extend the self in time? Future self-continuity rose only among older adults; a measure of continuity with one’s legacy or with future generations, not with one’s own self in five years, has not been tried.