Nik

Back to Psych 204 Projects 2013

Neural Correlates of Environmental Valuation: Background

The best method to capture the comprehensive value of environmental goods and services is a matter of contentious debate. Assessing environmental goods solely on the monetary potential of their natural resources as products and raw materials surely misses a large portion of the utility that humans derive from them. These benefits frequently fall outside the bounds of the marketplace, but an attempt to place them in an economic framework is necessary in order to arrive at some relative, scalable measure of their importance and value to the average citizen as public goods.

However, the majority of citizens are not directly involved in the purchase and management of public goods, necessitating the introduction of proxy measures such as surveys to derive these values (1). The contingent valuation (CV) survey method has been widely used by environmental economists – perhaps most visibly in its early days after the Exxon Valdez oil spill – as a survey method of assessing the nonmarket value of natural resources through respondents' willingness-to-pay (WTP) to prevent their loss or repair damages (2). Yet, the choices that individuals make in CV surveys are hypothetical, leading to problems with incentive compatibility and a disparity between their stated and revealed preferences (3). Moreover, survey responses in environmental valuation methods are subject to what Kahneman terms "affective valuation" (4), and are understood best not as true economic preferences, but as as expressions of emotion and attitudes. A large literature describes the seemingly irrational valuation decisions that plague environmental valuation techniques: "protest zeroes" where respondents refuse to put a price on the natural resource (5), stated WTP that is determined by the actions that damaged an environmental resource, rather than the value of the resource itself (1), and quantity insensitivity, where scaling the resource does not alter WTP (6). Jonathan Baron posits that environmental resources are often "protected values": highly resistant to economic tradeoffs, they elicit moral outrage when they are placed within a market framework (7).

To assess the degree of influence that affect has over WTP during environmental valuation, we designed a novel incentive-compatible donation task for functional MRI in order to identify the neural correlates of the emotional and cognitive responses involved in the decision process. The task elicited participants' willingness to pay to protect threatened state and national parks from proposed developmental land uses. We tested several predictions:
(i) Iconic natural landscapes would elicit activation in structures correlated with positive affect such as the ventral striatum
(ii) Destructive land uses would elicit activation in structures correlated with negative affect such as the anterior insula
(iii) WTP would be determined more, as Peter Diamond asserted, by response to the threatening land use than response to the resource itself (1)
(iv) Respondents with stronger pro-environmental attitudes, as measured by Dunlap's revised New Ecological Paradigm scale (8), would donate more frequently

MNI space

You can use subsections if you like. Below is an example of a retinotopic map. Or, to be precise, below will be an example of a retinotopic map once the image is uploaded. To add an image, simply put text like this inside double brackets 'MyFile.jpg | My figure caption'. When you save this text and click on the link, the wiki will ask you for the figure.

Below is another example of a reinotopic map in a different subject.
Figure 2

Once you upload the images, they look like this. Note that you can control many features of the images, like whether to show a thumbnail, and the display resolution.

Figure 3

MNI space

MNI is an abbreviation for Montreal Neurological Institute.

Methods

Selection of Stimuli

In order to assess affective response and subjective perceptions of our stimulus set, we performed a pair of online surveys via SurveyMonkey to determine which parks (n=36) and land uses (n=29) to use in the final task. Respondents of Survey 1 rated various photographs of parks on 7-point Likert scales in terms of affect (valence/arousal), familiarity, and iconicness. Respondents of Survey 2 rated uses on valence/arousal, moral outrage, and perceived destructiveness to the park. Additionally, respondents of Survey 2 were shown place/use combinations and asked to provide hypothetical WTP. This data allowed derivation of our stimulus set of 24 places and 24 uses. The 12 most iconic and 12 least iconic places and 12 most destructive and 12 most conservative uses were utilized in the study; parks and uses with multiple image variants in the survey used the image that scored closest to the mean in valence and arousal for that park or use in the final two counterbalanced stimulus sets.

WTP Donation Task

Participants were endowed with $24 at the beginning of the study, from which a randomized donation decision would be deducted. In a series of 72 trials, participants were first shown a park, then a proposed use for a quarter of the land (e.g., mining), and then asked if they would donate a specific amount (e.g., $15) to help avert this use (Figure 1). Parks and uses were chosen from our pilot survey data, and amount varied between $1 and $18, and was binned into low ($1-6)/medium ($7-12)/high ($13-18) categories. Participants were informed that the parks they saw were under threat of closure, a situation which could be averted by either: 1) sufficient donations or 2) selling off 25% of the park to a third-party buyer, who would put the land to a specified use (from mining to sustainable agriculture to rock-climbing camps).

At the study's conclusion, one trial was randomly selected to count for real; any donation was deducted from the subject's starting endowment and given to their choice of the California State Parks Foundation or National Parks. Subjects were instructed that since only one trial's results would be enforced, they should treat each decision as wholly independent of the others, and not attempt to strategically parcel out their endowment between multiple trials.

Questionnaire

Participants filled out a detailed questionnaire to obtain basic demographics, affect ratings for each place and use, familiarity with the locales, perceived destructiveness and public benefits of the uses, and whether they had visited each park before. In the fMRI population, these subjective ratings were used as regressors to model brain activation. Dunlap's revised NEP scale8 was also administered to evaluate pro-environmental attitudes. Participants were asked about their level of environmental concern and involvement in pro-environmental causes, how often they visited state and/or national parks annually, and how much they donated annually to environmental charities.

Participants

Twenty-nine healthy right-handed English-speaking adults participated in the study. Participants had no history of neurological or psychiatric disorders, and were not taking medication to interfere with fMRI. Seven subjects were excluded for excessive head motion and/or not following instructions.

MRI Acquisition and Analysis

Images were acquired with a 3.0 T General Electric MRI scanner using a thirty-two channel head coil. Forty-six 2.9 mm thick slices (in-plane resolution 2.9 mm isotropic, no gap, interleaved acquisition) extended axially from the mid-pons to the crown of the skull, providing whole-brain coverage and good spatial resolution of sub-cortical regions of interest (e.g., midbrain, NAcc). Whole-brain functional scans were acquired with a T2*-weighted gradient echo pulse sequence (TR = 2 s, TE = 24 ms, flip = 77º). High-resolution structural scans were acquired with a T1-weighted pulse sequence (TR = 7.2 ms, TE = 2.8 ms, flip = 12º) after functional scans, to facilitate their localization and co-registration.

Whole brain analyses were conducted using Analysis of Functional Neural Images (AFNI) software (9). For preprocessing, voxel time series were sync interpolated to correct for non-simultaneous slice acquisition within each volume, concatenated across runs, corrected for motion, slightly spatially smoothed to minimize effects of anatomical variability (FWHM = 4 mm), high-pass filtered (admitting frequencies with period <90 s), and normalized to percent signal change with respect to voxel means for the entire task. Visual inspection of motion correction estimates confirmed that no subject’s head moved more than 2.0 mm in any dimension from one volume acquisition to the next.

For whole brain analyses, regression models included eight regressors of no interest, six of which indexed residual motion, and two of which indexed cerebrospinal fluid intensity and white matter intensity (10). Regressors of interest orthogonally contrasted iconic parks versus noniconic parks, conservative versus destructive uses, low versus high requested donation amounts. Additionally, interactions were modeled at the time when all relevant information was first present (e.g., iconicness x destructiveness was modeled during presentation of the proposed land use, while destructiveness x amount was modeled during presentation of the donation amount). Prior to inclusion in the model, regressors of interest were convolved with a single gamma-variate function that modeled a canonical hemodynamic response (11). Maps of t-statistics for regressors of interest were transformed into Z-scores, coregistered with structural maps, spatially normalized by warping to Talairach space, and resampled as 2mm cubic voxels.

Results

Issues of Interest: Differing Subjective vs Population Maps

To examine generalizability of response as well as individual subjectivity, we created regressions based on the survey population's ratings (henceforth, "population" maps) and based on the subject's individual ratings.

In the process of analysis, we ran across something interesting (and mildly unsettling): we found relatively consistent results in the population maps for the first 18 subjects, which included ventral striatal activation for iconicness (Figure 2), bilateral insula activation for destructiveness of the land use (Figures 3 & 4), and interaction between the iconic x destructiveness in the MPFC (Figure 5).

Figure 2. Population maps at n=18, p<0.001 uncorrected, showing evidence of ventral striatal activation with increasing iconicness of the landscape.

Figure 3. Population maps at n=18, p<0.005 uncorrected, showing increased bilateral anterior insula activation with increasingly destructive land uses.

Figure 4. Population maps at n=18, p<0.001 uncorrected; anterior insula activation holds on right side.

Figure 5. Population maps at n=18, p<0.001 uncorrected; iconic x destructive interaction increases activation in the MPFC.

When four additional subjects were added to give us our final study population of 22, maps changed substantially. NAcc activation in response to iconicness became insignificant (Figure 6 shows it only holds until about p=0.006). Insula activation held on the left side (Figure 7). Iconic x Destructive interaction migrated higher up the MPFC to regions more dorsal than we would anticipate (Figure 9); we see the initial MPFC activation is still weakly evident at p=0.01 but does not survive at higher thresholds (Figure 8).

Figure 6. Population maps at n=22, p<0.01 uncorrected, weak NAcc activation in response to iconicness which does not survive higher thresholds.

Figure 7. Population maps at n=22, p<0.001 uncorrected, anterior insula activation remains strong on the left side in response to increasingly destructive land uses.

Figure 8. Population maps at n=22, p<0.01 uncorrected, weak response to interaction of iconicness x destructiveness of original MPFC region of activation which does not survive higher thresholds.

Figure 9. Population maps at n=22, p<0.001 uncorrected, MPFC activations remain strong in more dorsal regions not as frequently associated with affect integration in valuation.

Measuring retinotopic maps

Retinotopic maps were obtained in 5 subjects using Population Receptive Field mapping methods Dumoulin and Wandell (2008). These data were collected for another research project in the Wandell lab. We re-analyzed the data for this project, as described below.

Subjects

Subjects were 5 healthy volunteers.

MR acquisition

Data were obtained on a GE scanner. Et cetera.

MR Analysis

The MR data was analyzed using mrVista software tools.

Pre-processing

All data were slice-time corrected, motion corrected, and repeated scans were averaged together to create a single average scan for each subject. Et cetera.

PRF model fits

PRF models were fit with a 2-gaussian model.

MNI space

After a pRF model was solved for each subject, the model was trasnformed into MNI template space. This was done by first aligning the high resolution t1-weighted anatomical scan from each subject to an MNI template. Since the pRF model was coregistered to the t1-anatomical scan, the same alignment matrix could then be applied to the pRF model.
Once each pRF model was aligned to MNI space, 4 model parameters - x, y, sigma, and r^2 - were averaged across each of the 6 subjects in each voxel.

Et cetera.

Results - What you found

Retinotopic models in native space

Some text. Some analysis. Some figures.

Retinotopic models in individual subjects transformed into MNI space

Some text. Some analysis. Some figures.

Retinotopic models in group-averaged data on the MNI template brain

Some text. Some analysis. Some figures. Maybe some equations.

Equations

If you want to use equations, you can use the same formats that are use on wikipedia.
See wikimedia help on formulas for help.
This example of equation use is copied and pasted from wikipedia's article on the DFT.

The sequence of N complex numbers x₀, ..., x_N−1 is transformed into the sequence of N complex numbers X₀, ..., X_N−1 by the DFT according to the formula:

$X_k=\sum _{n=0}^{N-1}{x}_n{e}^{-\frac{2\pi i}{N}kn}k=0,\dots ,N-1$

where i is the imaginary unit and $e^{\frac{2\pi i}{N}}$ is a primitive N'th root of unity. (This expression can also be written in terms of a DFT matrix; when scaled appropriately it becomes a unitary matrix and the X_k can thus be viewed as coefficients of x in an orthonormal basis.)

The transform is sometimes denoted by the symbol $ℱ$, as in $𝐗=ℱ\left\{𝐱\right\}$ or $ℱ\left(𝐱\right)$ or $ℱ𝐱$.

The inverse discrete Fourier transform (IDFT) is given by

$x_n=\frac{1}{N}\sum _{k=0}^{N-1}{X}_k{e}^{\frac{2\pi i}{N}kn}n=0,\dots ,N-1.$

Retinotopic models in group-averaged data projected back into native space

Some text. Some analysis. Some figures.

Conclusions

Here is where you say what your results mean.

References - Resources and related work

References

Software

Appendix I - Code and Data

Code

File:CodeFile.zip

Data

zip file with my data

Appendix II - Work partition (if a group project)

Brian and Bob gave the lectures. Jon mucked around on the wiki.