best window tint business?

[Maverick]

Part 1

A few people have done 15% all round tints in NSW. Was just wondering if they ran into any problems with the police? The legal limit is 30% i think. Is it to dark?

Yes it is. Unlike some laws this one is based on scientific facts and these can't be refuted however the same brianiacs who fit this tint most likely also have illegal hid conversions and believe that they increased their visibility.

If they have an accident and the police attend the car at the very least will be ordered off the road and if the accident sees someone get injured you can bet the police will be looking long and hard at the illegal tinted windows and the role that they played in the accident. Not to mention the insurance company can refuse to pay out if they consider the tint played a part in the accident.

There are heaps of studies that have been performed that have proven that using dark tints on the road is stupid.

Visibility Through Tinted Automotive Glazing

(the ocr'ing on this documents sucks sorry but it's very relevant to the subject at hand)

Effects of car window tinting on visual performance: a
comparison of elderly and young drivers
N ICH O LA S R. B U R N S *, T ED N ETTELBECK , M ICH A EL W HITE and J A CQU ELIN E W ILLSON
Department of Psychology, University of Adelaide, South Australia 5005,
Australia
Keywords: Visual perception; Automobile driving; Perceptual masking; Aged;
Accidents; Tra c.
A major concern in allowing the tinting of car front side windows to 35% visible
light transmittance (VLT) is that tasks performed through these windows often
require the rapid detection of low-contrast, unilluminated targets. If the tinting
interferes with detection of targets then road safety may be compromized. Speed
of cognitive and visual processing declines with age; performance on backward
pattern masking tasks can indicate this slowing in processing speed. Two
experiments compared performance of the young and elderly adult on two
backward pattern masking tasks with levels of VLT from 100 to 20% . The ® rst
experiment found a decrement in performance for the elderly at 63% VLT and for
all participants at 20% VLT. The second experiment found a decrement in
performance for the elderly at 35% VLT. It was concluded that road safety may
be compromized if the front side windows of cars are tinted to 35% VLT.
1. Introduction
The focus of research and debate about acceptable levels for the tinting of car
windows, in terms of possible decrements in visual performan ce of drivers, has
shifted from concern m ainly with the front windshield and the prim ary vision area
(PVA) to concern with the front side windows of vehicles. This shift has occurred
because the parties to the debate, particularly road safety authorities and window
body-tinters and producers of window ® lm, are agreed, ® rst, that the PVA should
have a minimum visible light transmittance (VLT) of about 75% . (Australian Design
Rule ADR 8 / 00 speci® es 75% for the PVA, whereas the Federal M otor Vehicle
Safety Standard No. 205 [FM VSS 205] of the National Highway Tra c Safety
Administration [NHTSA] of the U SA speci® es 70% .) Second, it is agreed that for the
windows to the rear of the driver a minimum VLT of between 35 and 65% is
acceptable because of the very diŒ
erent demands of visual tasks likely to be
perform ed through these windows (Dain 1994).
For the front side windows of motor vehicles, the positions of the parties to the
debate are still opposed. Proponents for tinting windows argue that the minimum
VLT should be 35% for all windows except the windshield. They point, ® rst, to the
bene® ts of tinting windows to this VLT level (claimed to include glare reduction,
heat reduction, protection from harmful ultraviolet radiation, protection from
laceration, standardization in manufacture of car window glass, increased aesthetic
appeal of vehicles and enhan ced privacy for vehicle occupants); and, second, to the
fact that no detrimental eŒ
ects of tinting front side windows to 35% VLT have been
adequately demonstrated from road crash data. Road safety authorities, on the
other hand, argue, ® rst, that the claimed bene® ts of the tinting of windows m ay be
illusory and point to disbene® ts of such tinting (including the reduced ability of other
road users to see into the car through tinted windows and also to the concerns of
police in terms of o cer safety and suspect identi® cation). Second, road safety
authorities maintain that extant evidence from laboratory studies points to
decrements in visual performance associated with viewing out through tinted
windows, which are likely to compromise road safety and which are relevant to
driving tasks performed through the front side windows.
The main current argument against adopting the 35% VLT standard for front
side windows is that it would be detrimental to road safety through its eŒ
ect of
reducing visibility for the driver. Although there is an extensive literature on the
eŒ
ects of tinting on driver visual performance (reviewed by W hite 1992 , Jenkins
1994 , Sayer an d Traube 1994), many of the papers relate to the earlier debate
concerning the eŒ
ects of allowing the windshield to carry a light body-tint and are
not relevant to the current debate about darker ® lm tints on front side windows. The
earlier empirical studies were predominantly of night-time seeing distances using
low-contrast targets illum inated by car headlights. A num ber of other earlier papers
(discussed by Zwahlen and Schnell 1994 ) report the results of the mathematical
modelling of night-time seeing distances without presenting new empirical evidence.
These empirical an d modelling studies are obviously of little relevance to vision
through the front side windows, where the headlights have little or no eŒ
ect. Those
of the previous studies that have relevance to the current debate are now considered.
There are only six empirical studies, all of which have been reported in the past
10 years, that present evidence of potential relevance to the front side window debate
(Rom pe and Engel 1987, W akeley 1988 , 1992, Stackhouse and Hancock 1992 ,
Derkum 1993 , Freedman et al. 1993). In contrast with the earlier empirical studies,
these six studies were all laboratory-based. Participants performed a visual task
under conditions of low illumination that stimulated dusk or night driving.
Headlight illumination (real or simulated) was not used because objects detected
through the front side windows would not normally be illuminated by a car’ s
headlights.
These six studies included ® ve that used som e combination of measures of target
detection and response latency as the dependent variables and one (Derkum 1993)
that used measures of visual acuity. All of the studies included conditions with VLT
levels near 35% and illuminance levels that simulated twilight or night-time
conditions, or both. Three of the studies (Rompe and Engel 1987 , Derkum 1993 ,
Freedman et al. 1993) reported decrements in performance associated with
decreasing levels of VLT. Of the other three studies that did not, two (W akeley
1988 , 1992) have been criticized for using targets with contrast levels that were too
high to be aŒ
ected by reduced VL T levels (W hite 1992), and the third (Stackhouse
and Hancock 1992) was satis® ed with decrements in performance with 30% VLT
that were equivalent to a target detection failure rate of 23% .
The ecological validity of these studies is m ainly determined by the extent to
which the experimental procedures capture the essence of `conspicuity’ , and
particularly that aspect of conspicuity referred to by Cole and Jenkins (1980) as
sensory conspicuity (because the rapid detection and avoidance of an object in a
potential accident situation seems m ore dependent on actually seeing the target,
rather than attending to it because of its appraised meaning or relevance). There are
good reasons for believing that a measure of conspicuity in terms of the speed of the
response of the visual system might prove to be as valid and sensitive as any previous
measure. As part of normal driving, the driver is often required to make numerous,
fast ® xations to scan the visual ® eld adequately. There are some complex situations
where the driver’ s ab ility to scan the road environm ent adequately is tested to the
limit. Given that each ® xation takes longer under conditions of poor illumination, or
target contrast or both, drivers will sometimes be driving beyo nd the limits of their
scanning rate.
Backward pattern masking can be used to provide an estimate of the speed of the
visual system. If a target pattern is presented for a brief period, followed by a
meaningless visual `mask’ , the target will only be recognized if a su cient period (in
the vicinity of 50 ± 100 ms) is allowed to elapse between the onset of the target and
the onset of the mask. The minimum period between the presentation of the target
and mask that still allows the participant to recogn ize the target is known as `critical
stim ulus onset asynchrony’ (CSOA). The measurement of CSOA provides an index
of the speed of the visual system.
Backward masking CSOAs have been used in a number of applied contexts to
provide measures of the speed of processing of the visual system. It has been found, for
example, that CSOAs are signi® cantly longer for retarded participants (Nettelbeck
1985) and are signi® cantly lengthened by sm all quantities of alcohol (review ed by
M oskowitz and Robinson 198.
(goto part 2)

[Maverick]

part 2

Of more relevance to the driving task are the fndings
that CSOAs are longer for the elderly (summ arized by W hite 1996). This increase in
backward masking CSOAs for the elderly re¯ ects the more general decline in speed of
processing found in the elderly and is thought to be causally prior to the decline found
in performance by the elderly on measures of higher-order cognition (Nettelbeck and
Rabbitt 1992 , Kail and Salthouse 1994). Finally, backward masking CSO As are
probably longer under conditions of low illumination, and for low-contrast stimuli
(Hellige et al. 1979), although there is little relevant research on target and mask
energies. In the context of the current paper, it is worth noting that sex diŒ
erences are
not thought important for backward masking CSOAs (Nettelbeck 1987).

The experiments reported here compared backward masking CSOAs for the
elderly and young adult. The expectation was that the elderly would process visual
information more slowly (i.e. longer backward masking CSOAs). The backward
masking CSOAs were measured under conditions of high and low illum ination, and
with high- and low-contrast targets. It was expected that CSOAs would be longer for
low illumination and low-contrast target conditions. Finally, the CSOAs were
measured when the targets were viewed with several levels of VLT. In term s of
relevance to the debate about road safety, it was predicted that there would be an
interactive eΠof age, illumination, target contrast and VLT. That is, the elderly
ect
would be at a disadvantage in terms of their ab ilities to scan the visual ® eld
adequately; this disadvantage would be exacerbated by low illumination, low-
contrast targets and low VLT. This outcome would imply that tinting of front side
windows to 35% VLT may lead to an increase in road crashes involving elderly
drivers, or that elderly drivers may be restricted to daytime driving.
2. Experiment 1
The masking task used in this experiment arose from a model of comparative
judgem ent (Vickers 1970 , Vickers et al. 1972 ) and is termed `inspection time’ (IT).
There is a reliable negative correlation of about Ð 0.5 between CSOA on this task
(i.e. IT) and IQ and for this reason the task has been extensively studied (reviewed by
Nettelbeck 1987, K ranzler and Jensen 1989). Further, it has been demonstrated that
IT estimates are signi® cantly longer for the elderly (Nettelbeck and Rabbitt 1992). In
its most common form the task involves the presentation of a target stimulus
consisting of two vertical lines of markedly different length, joined at the top by a line. The target stim ulus is followed by a patterned mask and the participant
indicates on which side of the ® gure, left or right, the shorter line appeared.
The experiment reported here m easured IT under conditions that sim ulated day
or night driving illuminance (i.e. optimal versus marginal viewing conditions) and at
several levels of VLT. As noted above, the general decline in speed of processing (and
cognitive ability) found in the elderly means that IT estimates will be longer for
them. This, together with the interaction of this decline in speed of processing with
the effects of reduced visual information (low illumination, decreased VLT), leads to
the following hypotheses: (1) that elderly persons have longer IT estim ates than the
young adult; (2) that under marginal (`night-time’ ) viewing conditions IT estimates
will be longer; (3) that for these marginal viewing conditions IT estimates will
increase as VLT level decreases; and (4) this increase in IT estimate will be greater for
the elderly. Support for these hypotheses would imply that the speed of visual
processing in elderly drivers is comprom ised under marginal viewing conditions and
that this eΠis heightened when viewing with reduced VLT.
ect
2.1. M ethod
2.1.1. Participants: Thirty participants were recruited into two groups; one
comprised eight males and 10 females with a mean age of 70.6 (SD = 5.5) years,
and the other comprised three males and nine females with a mean age of 22.8
(SD = 2.4) years. The former group was recruited from the general community; 10
had education at the tertiary level and eight at the secondary level. Of the latter
group 10 were university students and two had education at the secondary level. It
was therefore considered unlikely that differences in IQ between groups would
influence the outcome. Potential participan ts underwent optometric testing (see
below) to preclude anyone with abnormal vision; all held a current driver’ s licence
but their driving habits were not ascertained. It is worth noting that, in South
Australia, drivers > 70 years are required to furnish a medical report annually and to
undertake practical driving tests at 75 years, 80 years and an nually from that time
on. The elderly sample here, then, had a mean age close to that at which the
legislative restrictions on elderly drivers begin to apply.
2.1.2. Optom etric testing: All potential participants attended an optometric clinic
for a standard optometric exam ination. Only those with vision, contrast sensitivity,
and ocular physiology deemed normal continued their participation in the
experiment.
2.1.3. Stimuli and tasks: Stimuli in the IT task were presented using a Gerbrands
G-1130 three-® eld tachistoscope, converted in-house to four ® elds. Internal
illumination of the tachistoscope was set at maximum intensity for all ® elds during
the optimal viewing condition (`day-time’ ) and was set at 85% for the marginal
(`twilight’ ) viewing condition. These settings produced lum inan ce levels of 4.5 and
1.1 cd / m2 respectively, measured at the tachistoscope eyepiece and with the
spotmeter pointed at the target approxim ately 0.75 m aw ay. These values are
recognized as somewhat unreliable but do dem onstrate the large luminance
diŒ
erence between the two conditions.
N eutral density ® lters were used in the tachistoscope eyepieces to produce three
levels of VLT with 63, 32 and 20% . These levels span a range of VLTs that begin at a
level somewhat lower than that deemed acceptable for the PVA of a motor vehicle,
include a condition close to that being argued for by the proponents of the 35% VLT
level for all windows except the windshield, and decrease to a level likely to be
attained by a window that has a 35% VLT level when new but that has deteriorated
in use (W hite 1994).
The stimulus cards for the tachistoscope were grey (with a colour density of 10
according to the Kodak Gray Scale) and the target and mask ® gures were
constructed on the cards with black precision slit tape. This means that the contrast
between the target and background was relatively low. The target ® gures were made
from 1 mm-wide tape and the mask ® gure was made from 4 mm-wide tape. The
alternative target ® gures consisted of two vertical lines joined at the top by a
horizontal line; the longer line subtended a visual angle of 2.6 8 and the shorter line
subtended a visual angle of 1.8 8 . The horizontal distance between the vertical lines
subtended a visual angle of 0.8 8 . The mask was similar in form to the target but was
symm etrical with both vertical lines subtending a visual angle of 3.4 8 (® gure 1). Each
trial began with the binocular presentation of one of the alternative targets (each of
the targets appeared equiprobably) in the centre of the visual ® eld. After the SOA
determ ined by the IT estimation algorithm (see below) the target was replaced by the
mask. The participant responded by pressing the corresponding key, left or right, on
a custom panel to indicate on which side the shorter line had appeared. If unable to
discriminate which ® gure had appeared the participant was required to guess. Two
seconds after the response, the next trial began.
(go to part 3)

[Maverick]

part 3

2.1.4. Design: Participants completed six sets of trials, corresponding to the six
possible combinations of luminance (two levels) and VLT (three levels). To
incorporate a number of practice trials at the beginning of each condition the initial
SOA was 450 ms. Additionally, the experimental design was partially balanced by
randomly allocating participants to one of four possible orders for completing the six
conditions, thereby controlling for any additional practice eŒ
ects. These four orders
of conditions were such that, for any participant, the VLT levels either increased or
decreased for both luminance conditions. The order of luminance conditions was
approximately balanced across participants, as were the alternative orders of VLT
levels.
2.1.5. Procedure: Participants attended one experimental session. Each was seated
before the apparatus in a small, arti® cially lit room, with the key panel in front. The
nature of the task, including the IT estimation procedure, was explained and
emphasis was placed on the requirement for accuracy in responding as opposed to
speed in responding. Participants were informed that they should take as much time
as they required to respond and that a new trial would not commence until after a
response was made. Before commencing the experimental trials, a series of practice
trials was provided to ensure participants understood what was required of them.
Those with prescriptions for visual correction were given practice both with and
without glasses; provided that they could correctly discriminate all practice trials
with ease, they were permitted to perform on the experim ental task without
prescriptions if this was more comfortable.
The entire sequence of stimulus presentations was autom ated by a computer
program , with brief rests between the VLT levels in each luminance condition and
for approximately 5 min between luminance conditions. The SOA on any trial was
determined according to an adaptive staircase algorithm (W etherill and Levitt 1965),
which was set to return an IT estimate at the 90% accuracy level; this required
approximately 100 trials for each condition.
2.2. Results
Repeated measures ANO VA showed a signi® cant three-way interaction between age
group, luminance level, and VLT level (Pillai’ s trace = 0.35, F(2,27) = 7.30,
p < 0.01). Figure 2 shows the means an d SD for IT, plotted for each condition. It
can be seen that decreasing VLT level was associated with increased IT for both age
groups in the low-lum inance viewing condition but that the elderly exhibited a more
marked increase. Polynomial contrasts demonstrated a linear trend for VLT level
that was signi® cantly diŒ
erent for the low- and high-luminance viewing conditions.
No diŒ
erence was apparent for the two groups in the high-luminance condition, but
the young and elderly exhibited diŒ
erent patterns of linearity in the low -luminance
condition (t (56) = 3.2, p < 0.01). Examination of ® gure 2 shows that mean IT was
longer for the low-luminance viewing condition irrespective of the VLT level and
that this eΠwas greater for the elderly (F(1,2 = 22.66, p < 0.001). Finally, the
ect
mean IT was longer for the elderly group in each condition (F(1,2 = 25.37 ,
p < 0.001). The results of the repeated measures ANOVA were therefore consistent
with the experimental hypotheses.
2.3. Discussion
The elderly exhibited longer mean ITs; this is consistent with the ® nding that speed
of processing decreases with increasing age. Under optim al viewing conditions VLT
level did not signi® cantly aŒ mean ITs of either age group. This ® nding supports
ect
the claim that tinting of car windows will not aΠdriver performance when viewing
ect
conditions are optimal and is therefore consistent with previous research that has
used target detection and response times as dependent variables. H ow ever, under
marginal viewing conditions VLT level did signi® cantly increase mean ITs and this
increase was signi® cantly greater for the elderly group. It can be concluded from this
that, under these marginal viewing conditions, elderly drivers were aŒ
levels as high as 63% (compared with performance under optim al viewing
conditions) and that performance deteriorated as VLT decreased. These results are
similar to those reported by Freedman et al. (1993). The young adults also showed a
deterioration in performance as VLT decreased from 63 to 20% .
This experiment did not include conditions equivalent to viewing through no
windows or through windows carrying only a very light body-tint. It is the
com parison of perform ance in these conditions to perform an ce with 35% VL T that
is critical in term s of relevance to road safety. There is also a question as to the
nature of the IT task, in that it has been shown to be vulnerable to the use of various
cues embedded in the task (W hite 1996 ) that may allow some participants to perform
at a level that is not determ ined strictly by the speed of the visual system. These
embedded cues may be a particular problem when the stimuli are presented
tachistoscopically. Experiment 2 was therefore designed to replicate and extend the
® ndings of this ® rst experiment by including conditions that directly compared both
a no window condition and a lightly tinted car window condition with a car window
tinted to 35% VLT. Second, an additional backward masking task that was
ostensibly free of embedded cues was included, thereby testing the generality of the
eŒ
ect of VL T on backward pattern masking CSOAs. Finally, all stimuli were
presented on a high performance computer monitor.
3. Experiment 2
Because experiment 1 had dem onstrated no decrement in performance with
decreasing VLT under optim al viewing conditions, this experiment included only
one level of illumination, which simulated twilight driving conditions. Two levels of
target contrast were employed, high and low. These levels sim ulated driving
conditions where headlight illumination is usually availa ble (that is, the front and
rear windows) and where no headlight illum ination is likely (that is, front side
windows) respectively. Three levels of VL T were used: in one condition there was no
window; in the other two conditions actual front side windows were used. One
window carried a light body-tint, the other had a ® lm applied to produce a VLT of
35% . Finally, in addition to the IT task used above, a second backward masking
task was designed. Theory (W hite 1996 ) and pilot studies suggested that this task
(known hereafter as the alphanumeric task because it involved recognition of
alphanumeric characters) was equivalent to the IT task but avoided the embedding
of movement cues within the task.
W hite (1996) has argued that the relevant criterion for designing pattern masking
tasks is that the targets are clearly distinguishable from the mask; the num bers of
targets or the relationships between targets are irrelevant. On these grounds the ® rst
hypothesis was therefore that the two tasks would be equivalent except that the
alphanumeric task may provide a `purer’ measure of speed of visual processing.
Nevertheless, the two tasks should produce the sam e pattern of results to diŒ
erent
levels of VLT or target contrast.
Speci® c hypotheses were: (1) compared with young adults, elderly adults have
longer ITs (in the case of the alphanumeric task the more general term CSOA will be
employed); (2) IT (or CSOA) will not be signi® cantly diŒ
erent for the no-window
and light body-tint window conditions, irrespective of the target contrast; (3) for the
35% VLT condition, IT (or CSOA) will be signi® cantly longer than for the 100 and
81.3% VL T conditions; and (4) this increase will be greater for the elderly group and
will be greater for the low-contrast target condition than for the high-contrast target
condition. Support for these hypotheses would imply that 35% VLT level would
ect all drivers when viewing through the front side windows of
vehicles under night-time conditions; this eŒ
ect would be exacerbated for elderly
drivers because of the slower processing speed found in the elderly.
3.1.1. Participants: Twenty-six participants were recruited into two groups; one
comprised nine males and four females with a mean age of 73.9 (SD = 2.4) years,
and the other comprised four males and nine females with a mean age of 24.0
(SD = 2.2) years. The former group was recruited from the general community;
® ve had education at the tertiary level and eight at the upper secondary level. All
of the latter group were university students. It was again considered unlikely that
erences in IQ between groups would in¯ uence the outcome. Potential
participants underwent optometric testing (see below) to preclude any with
abnormal vision; all held a current driver’ s licence and drove a vehicle at least
weekly. Elderly participants were paid A$30 on completion of their participation in
3.1.2. Optom etric testing: All potential participants attended the sam e optometric
clinic as in experiment 1. Again, a standard optometric examination was performed.
(goto part 4)

[Maverick]

part 4

Only those participants with vision, contrast sensitivity and ocular physiology
deemed normal continued their participation in the experiment.
3.1.3. Car windows : Two identical, new front side windows were purchased from a
local spare parts supplier. For one, an automotive window tinting retailer applied a
tint ® lm. VLT levels were measured in accordance with the Code of Practice for
Surface Film s for M otor Vehicle W indows (South Australian Departm ent of
Transport, April 1992). The windows as purchased had a VLT of 81.3% ; after the
application of the ® lm the tinted window had a VLT of 35.1% .
3.1.4. Stimuli and tasks: All stimuli were presented on a high-performance monitor
(with a vertical refresh rate of 120 Hz) under the control of a PC running at 90 M Hz
and ® tted with a video accelerator card. The tim ing of stimuli was controlled by use
of the vertical retrace ¯ ag of the video card and the stimuli were redrawn during the
vertical retrace; this meant that the stimuli were presented for durations that were
multiples of 8.3 ms. To manipulate target contrast the stim uli were drawn in two
com binations. The ® rst combination used white targets and masks presented on a
black background; this was the high-contrast condition. For the low-contrast
condition light-grey targets and masks appeared on a dark-grey background.
The IT task was very sim ilar to that used in experiment 1. The alternative target
stim uli consisted of two vertical lines joined at the top by a horizontal line; the longer
line subtended a visual an gle of 0.6 8 and the shorter line subtended a visual angle of
1.2 8 . The horizontal distance between the vertical lines subtended a visual angle of
0.6 8 . The mask consisted of the same horizontal line as in the target ® gure and two
vertical lines which subtended a visual angle of 1.8 8 . These lines were the same
thickness as the target lines and overwrote the target lines. The alphanumeric task
consisted of the presentation, on each trial, of one of eight segm ented alphanumeric
characters: `2’ , `3’ , `5’ , `7’ , `F’ , `H’ , `U’ and `Y’ . Each element of the alphanumeric
characters subtended a visual angle of 0.5 8 . The mask consisted of a segmented
matrix consisting of 31 elements each subtending a visual angle of 0.5 8 (® gure 3). The
central elem ents of this mask, in various combinations, constituted the target ® gures.
Each trial began with the presentation of two small circles, one above and one
below the position where the target would appear; after 523 ms the target was drawn
and rem ained on the screen for the SOA determined by the estim ation algorithm, the
initial SOA being 315 ms. Each of the targets appeared equiprobably. After the SOA
the mask was drawn and remained on the screen for 374 ms. The participant
responded verbally as to which side the shorter line had appeared (for the IT task),
or nominated which of the characters had appeared (for the alphanumeric task). If
unable to discriminate which target had appeared the participant was required to
guess; the experimenter pressed the corresponding button on a custom response
board. Two seconds after the response, the next trial was presented.
3.1.5. Design: There were two exp erimental sessions. At one session conditions
involving the IT task were com pleted and at the other conditions involving the
alphanumeric task were completed; the order of sessions was balanced across
participants. W ithin a session there were six experimental conditions corresponding
to two levels of target contrast (high versus low) an d three levels of tinting (100, 81.3
and 35.1% VLT). The tasks involving one of the contrast levels were completed in
the order 100, 81.3 and 35.1% VLT. The tasks with the other level of contrast were
then completed in the same order; the order of contrast was balanced across
participants. Additionally, the 100% VLT condition, with the high-contrast target,
was repeated at the end of the experimental session. This order of VLT conditions
meant that the practice eŒ
ect for each task operated against the hypothesis of
decrement in perform ance (i.e. increase in IT or CSOA) with decrease in VL T level.
3.1.6. Procedure: The participant was seated in a dimly lit room (vertical
illumination at eye-level was 2.2 lx) at 0.75 m from the computer monitor.
Spectacles, specially prescribed to correct vision for this viewing distance were worn
by 10 of the elderly participants. For the conditions requiring viewing through one of
the windows, the window was placed between the participant and the computer
monitor at 0.25 m from the monitor.
The task to be performed at the experimental session was explained using
diagram s and presentations of unmasked stimuli on the monitor. U nm asked trials
were then presented with a duration of 830 ms; all participants practiced until
performance was error-free for 20 consecutive trials. To familiarize the participant
with the low-contrast target, viewed through the 35.1% VLT window, 10 unmasked
trials with a duration of 830 ms were presented. N ext, the participant was required
to perform to a criterion level for the low-contrast target viewed through the 35.1%
VLT window; 15 of 16 unmasked trials with a duration of 33 ms were required to be
correctly discriminated. All participants met this criterion. These explanations,
practice and criterion procedures required approxim ately 30 min, which was
considered su cient time for dark adaptation (Davson 1990 : 219 ± 220).
Next, the estimation procedure was explained with emphasis placed on the
requirem ent for accuracy of responding as opposed to speed of responding. Before
the ® rst experim ental condition comm enced three sets of practice trials for this
condition were presented. These consisted of 10 trials each at SOA s of 830, 415 and
315 ms; perfect performance was required at the ® rst two SOAs and nine out of 10
(goto part 5)

[Maverick]

part 5

correct trials were required at the last SOA before the experimental conditions
began. In this experiment the adaptive staircase algorithm was set to return an IT
estimate (or CSOA) at the 80% accuracy level.
3.2. Results
The results for the two pattern masking tasks were analysed separately. Figure 4
shows the mean IT m easures plotted for each condition.
Repeated measures ANOVA determ ined that the three-way interaction between
age group, target contrast and VLT level was not signi® cant (Pillai’ s trace = 0.12,
F(2,23) = 1.66, p = 0.21). The pattern of the results was, however, similar to that of
experiment 1 and the non-signi® cance of the three-way interaction appears to be due
to the great variability in performance on the IT task of the elderly sample. It can be
seen in ® gure 4 that, for the elderly group, the SDs were approxim ately half the
magnitude of the mean IT estimates but for the young group they were only one-
third of the magnitude of the mean IT estimates.
Examination of ® gure 4 con® rms that the elderly had longer IT estimates than
the young (F(1,24) = 12.55, p < 0.05). The interaction between target contrast and
age group was signi® cant (F(1,24) = 10.03, p < 0.05); IT estimates were longer in the
low-contrast target condition for the elderly. It can be seen that, as expected, there
was little diŒ
erence between the IT estimates for the 100 and 81.3% VL T conditions
in either contrast condition for either age group (t(24) = 0.84, p = 0.41). Conse-
quently, tests of the eΠof 35.1% VLT on IT were m ade against the mean of the
ect
100 and 81.3% VL T conditions. The interaction between age group and the eΠof
ect
35.1% VLT was signi® cant (t(24) = 3.89, p < 0.001), an d the interaction between
contrast and the eŒ of 35.1% VLT was signi® cant (t(24) = 2.58, p < 0.05). It can
ect
be concluded that the elderly had increased IT estimates with 35.1% VLT and that
this increase was greater for the low-contrast target condition. As was the case for
the overall three-way interaction (see above), the interaction between target contrast
eŒ
ect, the eŒ
ect of 35.1% VL T and age group was not signi® cant (t(24) = 1.34,
p = 0.19). The results for this task, then, support the experimental hyp otheses and
are consistent with the outcome of experiment 1.
Contrary to our hyp othesis, the outcome for the alphanumeric task was diŒ
erent
from that of the IT task. Figure 5 shows the m ean CSO A measures plotted for each
condition.
Examination of ® gure 5 con® rms that the elderly had longer CSO A estimates
than the young (F(1,24) = 13.23 , p < 0.05). U nlike the IT task, however, there was a
signi® cant diŒ
erence between the 100 and 81.3% VLT conditions (t(24) = 2.65,
p < 0.05); this may re¯ ect a practice eŒ
ect. The extent of any practice eΠcan be
ect
gauged by comparing performance on the high-contrast target with 100% VLT
condition with the estim ate for the sam e condition repeated at the end of the
experiment. For approximately half of the participants this was the ® rst experimental
condition com pleted and for the others it was the fourth experimental condition.
Therefore, any im provement in performance found in the repeated performance of
this condition will underestimate the extent of any practice eŒ
ect. The m ean CSOA
for the experimental condition for the young sample was 44 ms (SD = 10) and for
the repeat of the condition 37 ms (SD = 9). This was a signi® cant improvement
(t(12) = 4.48, p < 0.05). For the elderly, the mean CSOAs were 60 m s (SD = 14) and
51 ms (SD = 12) respectively (t(12) = 3.07, p < 0.05). The interaction between target
contrast and age group was not signi® cant (F(1,24) = 3.94, p = 0.06). Therefore, to
test the hypotheses with respect to 35.1% VLT the ANOVA was restricted to two
levels of VLT, 83.1 and 35.1% . The interaction between age group and the eΠof
ect
35.1% VLT was signi® cant (F(1,24) = 5.63, p < 0.05); examination of ® gure 5 reveals
that this was due to an increase in CSO A for the elderly. There were no other
signi® cant eŒ
ects. Nevertheless, this task again demonstrated an eŒ
ect, on the
elderly, of decreasing VL T to 35% . This ® nding then is consistent with that found
for the IT task in this experiment and in experiment 1. The diŒ
erence between the
overall outcome on this task compared with the IT task, however, suggests that the
assumption that this and the IT task are qualitatively the same was mistaken; the
implications of this are discussed below.
3.3. Discussion
The ® nding from experiment 1 that backward pattern masking CSOAs (including
IT) were longer for the elderly was replicated. In experiment 2, however, it was found
that for the IT task this decrement in performance for the elderly was not consistent
across participants. Those participants found to have longer IT estimates under
optim al viewing conditions showed the greatest decrement in performance for the
worst-case viewing conditions. This ® nding may have practical implications in that
not all elderly drivers will be aŒ
ected to the same extent by decreased VLT levels.
That some will be quite seriously aŒ
ected means that legislation concerning
perm issible VLT levels for car windows needs to be framed to take account of this
group.
It was found that there was no diŒ
erence between viewing with 100% VLT or
81.3% VLT, except for an improvement in performance on the alphanumeric task
for the 81.3% VLT condition attributable to practice. The implication of this is that
the use of windows carrying a light body-tint poses no increased risk of road
accidents even in marginal viewing conditions. On the other hand, for both the IT
task and the alphanumeric task the results showed that the elderly are detrimentally
aŒ
ected by reduction in VL T level to 35% ; the IT or CSOA estimates were longer for
the elderly with 35% VLT and this eΠwas greater for low-contrast targets. The
ect
implication of this is that, for the conditions most likely to apply when viewing
through the front side windows of a vehicle at night-tim e, the elderly are
com prom ised in their ability to scan the road environm ent adequately.
The diŒ
erences found between the IT task an d the alphanumeric task are in some
respects puzzling. W hite (1996) explained that according to the integration theory of
backward pattern masking the `relevant criterion for a simple sequential ® gural
discrimination is that the targets, however many they might be, are clearly
distinguishable from the mask’ (p. 356). It was on this basis that the alphanumeric
task used here was designed; pilot studies had shown that performance with this task
and the IT task were equivalent. The results here, then, were contrary to our
expectation. The explanation for this appears to be two-fold. In the pilot study
referred to the participants were highly practiced at the IT task and it is possible that
rather than making a judgement as to which of two lines was shorter they were
(goto part 6)

[Maverick]

part 6 (and final)

recognizing the ® gure as a whole. This is precisely what the alphanumeric task
involves. Rather than making a discrimination or judgement about the target ® gure,
what is actually required is the recognition of the target. Given that our participants
were all highly familiar with the letters and num bers used as the targets, but naive
with respect to the IT task, then this may account for the diŒ
erence in the pattern of
the results for the two tasks. That is, rather than using highly overlearned characters
as targets it would have been better to design a task with more targets to be
discriminated but that were unfamiliar to the participants.
There are practical implications arising from this unexpected ® nding. Given that
the major concern with respect to the level of tinting for front side windows is with
whether lowered VLTs will increase the likelihood of road crashes due to a failure to
detect the presence of low-contrast targets under night-time viewing conditions,
then, our results suggest that concern is warranted. The IT task requires the
detection of a small diŒ
erence in the length of two lines, not the recognition of an
overlearned character. Similarly, the detection of the presence of a cyclist or
pedestrian through the front side window on a rainy night at an ill-lit intersection
requires the detection of a small diŒ
erence in ® gure-background contrast or of a
¯ icker of movem ent. The IT task may have better captured the essence of sensory
conspicuity and therefore the results for the task may be m ore ecologically valid.
4. General discussion and conclusions
As noted above, this study was the ® rst to use backward pattern m asking tasks to
index the eŒ
ect of reduced VLT on visual performance. Experiment 1 clearly
demonstrated that there was a decrement in performance associated with decreased
VLT level over the range 62 ± 20% . This ® nding was consistent with those of earlier
studies (Rompe and Engel 1987 , Derkum 1993 , Freedm an et al. 1993). Experiment 2
attempted to determine whether this decrement in performance applied for
conditions that sim ulated the driving tasks associated with driving in marginal
conditions and speci® cally those tasks performed through the front side window.
The outcome was that 35% VLT, the level being proposed as the standard for all
windows except the windshield, deleteriously aŒ
ected the elderly. It was noted that
this eŒ
ect was more pronounced for those of the elderly whose speed of visual
processing showed most senescent decline. However, all participants in this study
had been found to have normal vision and ocular physiology and all were licenced
drivers.
As noted in the discussion of experiment 2, the IT task may have provided some
ecological validity in assessing the eŒ
ects of reduced VLT on elderly drivers. Three
earlier laboratory based studies (Rompe and Engel 1987 , Derkum 1993 , Freedman et
al. 1993) also found perform ance decrements as VLT was decreased. However, an
empirical study in the ® eld may be required to provide de® nitive evidence on the
eŒ
ects found in the experiments reported here.
One issue that was raised by some of the elderly participants in these experiments
is worth commenting on; often elderly drivers are aw are of their lim itations and limit
their driving accordingly. For example, the elderly are believed to reduce their rate of
everyday, night and peak-hour driving (Ernszt and O’ Connor 198. Cognisance of
this should be tempered by the expectation that, in W estern societies, the num bers of
elderly drivers is increasing rapidly (in line with the ageing of these populations).
On the basis of the experiments reported here, the conclusion is drawn that the
level of tinting for front side windows should not be reduced to 35% VLT. A
conservative level would be the same as that for the PVA of the windshield. The
results found here suggest that such a level poses no increased risk of road crashes
due to decrem ent in visual performance.
Acknowledgements
W e are grateful to Caroline M iller and Sally G oodwin for recruitment of the
participants and collection of the data for the second experiment. Our thanks to
Andrew Gri ths for the optometric examinations and his input to the research. The
® rst exp eriment form ed part of the requirements for an Honours degree in Psychology
made by J.W ., supervised by M .W . The second experiment was supported by a grant
from the Australian Federal O ce of Road Safety to Ted N ettelbeck.

THE END

[Tux]

There you go. Pretty obvious really.

[DracZ]

Anyone got cliff notes? 2 pages for window tint?

[Corey_R]

The cliff notes are, if you run below 35% tint on your front driver and passenger windows, you're an idiot, because unlike many other 'discussions' which are based on opinion, this one is based on scientific fact!

As Maverick stated, this is one of those laws where there is conclusive scientific evidence to back it up.

[tdiwabbit]

Those that live east of Melbourne I can recommend very highly, John from Ringwood, he uses scotch 3m carbon tint and comes up looking very nice, it is not metalic. He did two of my new mazdas and friends golf few years ago, getting new cars and will be using him again. Twilight auto tint, ringwood. Lifetime warranty against colur change (purpel), bubbling, demetalising.

[emuexport]

Can anyone recommend a perth tinter?

[JIMMYP]

Does anyone know what type of tint matches the rear privacy glass on the MKVI Golf? There was another discussion thread about this, but doesn't seem to have resolved the issue. Can anyone help?

Also, is it possible to get just plain UV tinting (i.e. no colour/darkness change)?

Thanks alot!

[nau]

Also, is it possible to get just plain UV tinting (i.e. no colour/darkness change)?

Yes there is one, ask in the shop cost pretty much the same as all others
from what I saw on a demo it adds a little bit of reflection when you look on a angle but other thn that got no color what so ever

[Corey_R]

Yes there is one, ask in the shop cost pretty much the same as all others
from what I saw on a demo it adds a little bit of reflection when you look on a angle but other thn that got no color what so ever

Are you sure it costs no more than the rest?
The 'clear' tints use real gold as their metallic layer to block heat etc. Because of this, they're at least twice the cost of a other tints which use steel and other similar metals.

Unless I'm getting confused and you're literally talking about something which JUST blocks UV and does NOT block heat.

[nau]

Are you sure it costs no more than the rest?

dont quote me on that but when I was doing my tint last week I asked about prices for everything they had on display and from memory they were all more or less the same (+\-$50-100) difference except some protection tint (that makes your windows stronger) but
give TintACar a call for example and chk