Hypothesis testing
H5 – Selective Displacement (SDI) and Indoor–Natural Trade-off (CLR)
H5a Selective Displacement Index (SDA) Ordering
H1: SDI(TDM) > SDI(Built) > SDI(Natural), where SDI = β̂LB − β̂SA (positive = shift toward belonging).
H0: No ordering in SDI across predictors.
H5b CLR indoor – Natural Trade off
H1: Built and Natural are negatively associated within the DA composition.
H0: No association.
Question:
Do screen/indoor contexts shift need salience toward belonging over self-actualisation more than other cues (H5a), and is there a systematic trade-off between indoor/built and natural content within images (H5b)?
Data
Data subset:
All 10 Developmental Affordance (DA) groups and all 6 Maslow need levels from the full corpus. No temporal filters. Inclusion rule: images were retained if they had at least one DA or at least one Maslow category present after merge and cleaning (deduplicated at image × category).
For H5a – Selective Displacement Index for a predictor X as SDI(X) = (effect of X on Love and Belonging) − (effect of X on Self-Actualisation).
-
- A positive SDI means X shifts need salience toward belonging and away from self-actualisation.
For H5b – Model within-image shares using a centered log-ratio (CLR) transform and regress clr(Natural) on clr(Built).
-
- A negative slope means that as Built increases, Natural decreases within the same image—i.e., a trade-off.
Variables:
-
- Predictors (DA groups): Technology & Digital Media (TDM), Built Structures & Indoor/Urban Spaces (Built), Natural Landscapes & Phenomena (Natural), Humans.
- Outcomes (Maslow): Love & Belonging (LB), Self-Actualisation (SA).
- Controls: Category fixed effects (FEs); Humans included alongside TDM/Built/Natural in relevant models.
- Multiple testing: Benjamini–Hochberg FDR where p-values are available.
Normalisation & Weighting: See WWeighted Calculation Methods reference page.
N: 6,7890
Methods
H5a — Selective Displacement Index (SDI)
The Selective Displacement Index for a predictor X is defined as SDI(X) = (effect of X on Love and Belonging) − (effect of X on Self-Actualisation). A positive SDI means X shifts need salience toward belonging and away from self-actualisation.
-
- Weighted SDI: computed from the weighted proportion regressions (WLS) for LB and SA.
- Raw SDI: computed from the binary (logit) LB and SA models.
- Uncertainty for raw SDI: nonparametric bootstrap (1,000 resamples, blocked by category) to obtain 95% confidence intervals and ordering probabilities (e.g., P[TDM SDI > Built SDI]).
H5b — Indoor–Natural trade-off
Test whether Built and Natural are inversely related within the image composition.
Weighted (primary; compositional):
-
- CLR–OLS: regress clr(Natural) on clr(Built); a negative slope indicates a trade-off;
- Weighted Spearman on CLR values (monotone association).
Raw (presence/absence):
-
- Phi coefficient of Built_present × Natural_present, with chi-square test;
- Category-residualised Pearson correlation (demean each variable within category);
- Logit with category fixed effects: Natural_present ~ Built_present + category dummies.
Results
H5a Selective Displacement Index (SDI) ordering
| Predictor | Weighted SDI (mean) | 95% CI (weighted) | Raw SDI (mean) | 95% CI (raw) |
|---|---|---|---|---|
| Technology & Digital Media | −0.005 | [−0.118, 0.084] | 0.233 | [−0.124, 0.556] |
| Built (Indoor/Urban) | −0.209 | [−0.260, −0.155] | 0.673 | [0.389, 0.951] |
| Natural (Landscapes) | −0.656 | [−0.748, −0.560] | −2.264 | [−2.725, −1.823] |
Results – H5b — Built gradient across need levels (downward)
| Specification | Model/Test | Estimate | 95% CI lo | 95% CI hi | p-value | Extra |
| Weighted (composition) | CLR–OLS: clr(Natural) ~ clr(Built) | -0.126 | -0.148 | -0.105 | 0.0 | Share trade-off |
| Raw (co-occurrence) | Phi (Built×Natural, 0/1) | -0.004 | -0.028 | 0.019 | 0.714 | Co-occurrence |
| Raw (co-occurrence, FE-adj) | Pearson (demeaned by category) | 0.044 | 0.02 | 0.068 | 0.0 | Co-occurrence (FE) |
| Raw (FE-logit) | Natural_present ~ Built_present + FEs (β / OR) | 0.237 | 0.058 | 0.417 | 0.01 | OR=1.268 [1.060, 1.517] |
Combined Interpretation
Selective displacement (H5a): The nature vs digital/indoor split is robust.
-
- In weighted shares, Natural strongly tilts toward Self-Actualisation (negative SDI), Built is modestly negative, and TDM is ~0.
- In raw presence, Built and TDM tilt toward Belonging, Natural toward Self-Actualisation. The exact rank of TDM vs Built is measurement-sensitive (shares vs presence).
Indoor–natural trade-off (H5b): In the compositional lens (CLR on shares), Built and Natural trade off strongly; raw co-occurrence tests answer a different question and need not be negative.
Concluion
H5a — Reject H₀ (directional displacement). Natural aligns with Self-Actualisation, while Digital/Indoor aligns with Belonging. The strict TDM vs Built rank depends on measurement (shares vs presence), but the nature vs digital/indoor split is robust.
H5b — Reject H₀ (indoor–natural trade-off). In the compositional view of shares (CLR), Built and Natural move in opposite directions; raw co-occurrence tests answer a different question and don’t overturn this trade-off.