W2 : Basic statistical understanding in sensory science and psychophysics Flashcards
What does psychophysics investigate?
Pyschophysics quantitatively investigates the relationship between physical stimuli and the sensations and perceptions they produce.
Psycho : related to brain / thinking -> how brain intepret and sense/perceive stuff
Physical : physical stimuli
Psychophysical theories can be expressed in the form of mathematical functions, where S = f(I).
What is S, f and I?
S = f(I), where
<br></br>
S : sensation / perception
f : function of
I : intensity of stimuli
What is another name for absolute threshold and what is its definition?
Absolute threshold : aka detection threshold.
<br></br>
It is the point at which a substance can be differentiated from the background.
What is the general definition of a threshold. What 4 different thresholds are there in sensory science?
A threshold is a concentration range below which an odor / taste of a substance will not be detectable under any practical circumstances;; and above which individuals with a normal sense of smell and taste would readily detect the presence of the substance
<br></br>
Types of thresholds:
1. Absolute threshold (detection threshold)
2. Recognition threshold
3. Difference threshold
4. Terminal threshold
What is the definition of recognition threshold?
The point at which substance is correctly recognised and identified.
What is another name for difference threshold and what is its definition?
Difference threshold : aka Just Noticeable Difference (JND)
<br></br>
The point at which the change of concentration / intensity is noted. (i.e. how much you must increase the concentration by in order for one to perceive a difference)
What is the definition of terminal threshold?
The point at which no further intensity increase (perception) is found with increasing concentration.
What does Weber’s Law state?
Weber’s law states that the just noticeable difference (JND/difference threshold) between 2 stimuli is a function of the magnitude of the original stimulus.
- i.e. the larger the intensity of the original stimulus, the larger the JND needs to be to perceive the difference.
What does Fechner’s Law state?
Fechner’s Law states that the subjective sensation is proportional to the logarithm of the stimulus intensity.
- i.e. as stimulus increases, the sensation / perception increases non-linearly (aka logarithmically)
What does Steven’s power law state? What is the equation?
Steven’s power law states that subjective intensity (perceived intensity) is a power function of physical intensity (actual concentration / intensity of stimuli)
<br></br>
S = k In, where
<br></br>
S = Sensation / perceived intensity ; I = intensity of stimulus ; n = exponent of the power function ; k = constant that is specific to sense measured and type of stimulus
Steven’s power law : S = k In
What does it mean when n > 1 and when n < 1 ?
Steven’s power law : S = k In
<br></br>
When n > 1 : the perceived sensation grows faster than the stimulus –> a slight increase in stimulus leads to an even greater increase in perception / sensation
<br></br>
When n < 1 : The sensation grows slower than stimulus –> i.e. for the 1 unit increase in stimulus, the perception intensity increases less
The Biedler model proposed that taste response can be approximated by mass action law.
Looking at the graph in the notes, the curve of taste response (R/Rmax) against concentration follows a sigmoidal curve. Explain the shape of the curve.
At lower concentrations, taste receptors are not saturated yet, thus as concentration increases, increase in taste perception is much faster.
<br></br>
At higher concentrations, as taste receptors become more saturated with molecules, the rate of increase in taste perception decreases (as shown by the decreasing gradient), and eventually plateaus when taste receptors become fully saturated.
What are the 2 methods in determining thresholds?
- Method of limits
- Method of constant stimuli
How is method of limits conducted?
The ohysical stimulus is changed by successive discrete steps until a change in response is noted.
- e.g. successively increasing / decreasing concentration of sucrose solution
The method of limits can only apply for detection / absolute threshold. True or False?
False. The method of limits can also apply for recognition threshold.
<br></br>
Just differ in the questions asked.
<br></br>
For detection threshold, qn asked will be “can you perceive anything ?” – e.g. sweetness detection
<br></br>
For recognition threshold (e.g. recognising taste / aroma) : “What flavour / aroma is this?” –> then keep increasing concentration until they can say the correct ans.
How does the method of stimuli work in determining thresholds?
A reference / control is given, and the test stimulus is compared against this reference / control.
For example, I want to determine the detection threshold for sucrose solution, ranging from 1 - 10 ppb. How can I use the method of limits to determine the detection threshold?
- Present panellists with increasing concentrations of sucrose solutions, starting from 1 ppb.
- Keep subsequently increasing the concentration by serving samples 1 by 1 and ask “Do you perceive anything?”
- Once the panellist states that they perceived a sensation (i.e. sweetness), the concentration at which they perceived it is their detection threshold.
For example, I want to determine the detection threshold for sucrose solution, ranging from 1 - 10 ppb. How can I use the method of constant stimuli to determine the detection threshold?
- Provide a reference sample (e.g. concentration = 6ppm).
- Serve all the samples at once in a randomised order.
- Instruct panellists to taste the reference, and taste each sample in order. Ask panellists to respond to each test sample as “sweeter than” or “less sweet” than the standard.
- Obtain a curve of frequency / % judged “sweeter” than standard against sucrose concentration.
What is nominal data and its purpose?
Categorical data.
Purpose is to count the frequency of each category / number.
Since nominal data is categorical data, it must be in alphabets and cannot be in numbers. True or False?
False. Numbers can be used to label categories, just that the categories will not have mathematical meaning and are only used as labels
What is ordinal data and its purpose in sensory science?
Ordinal data (“order”) is data that follows a ranking order / relative order. Its purpose is to assign numbers in order to rank products with regard to some sensory properties.
Scaling methods : Interval scale
What is an interval scale? Interval scales allows for the comparison of?
It is a line marked with markings (1 mark = 1 response), and there is an equal spacing between each marking. It allows for the comparing degrees of differences.
What are the 2 kinds of interval scales in sensory science?
1) Linear scale - to do with numbers (e.g. line scale with number markings 1-10, and instruction is “please mark the sweetness intensity of sample” –> panellist can mark anywhere on the line, the markings just acts as a guide.
<br></br>
2) Category scale – to do with categories (e.g. line scale with 3 markings : weak, moderate and strong. instruction is “please circle the sweetness intensity of sample”) –> panellists choose 1 category and circle
Sensory data type : Ratio (relative sizes). What is a “ratio” data type?
A number indicated by panellists how many times a presented stimulus is stronger than a reference stimulus.
- not used often
Category scales
What kinds of tests can category scales be employed in?
- Hedonic / affective tests - indicate degree of liking (by consumers)
- Descriptive test - rating perceived intensity of a stimulus (by trained panellists)
What are the 4 kinds of category scales?
- Integer scale (numerical)
- Verbal scale (labelled)
- Check-box scale
- Facial scale (graphic)
In a category scale, what is the benefit of a 9 point scale over a 3 point scale?
With more discrete points, there are more divisions to provide you with more sensitive information on how products differ in specific attributes quantitatively. It also generates more variability in responses, increasing power of statistical analysis, where it is easier to tell significant differences between products.
<br></br>
- i.e. more graduations = allowing respondents to express finer distinctions in their preferences or perceptions. This enhances the ability to detect subtle differences between samples that a 3-point scale might miss.
Line marking scales are often used in descriptive analysis involving trained panellists. How does line marking scale work?
The line marking scale is labelled with start and end points (and sometimes, middle point), and trained panellists will mark the intensity of attribute anywhere along the line.
<br></br>
- The line could be 10cm long (representing intensity of 0 - 10)
- After the trained panellists mark their marking, use ruler to measure the distance of marking from start point , and that will be the intensity rating.
Why is it important to not put too many graduations on a line marking scale?
If there are too many points labelled on the line, it will resemble a category scale and panellists will be inclined to just mark on one of the graduations –> does not capture the true perception of panellist because the percived intensities are more like in discrete categories instead of a continuum.
When is ranking mainly used?
Hedonic / affective testing to rank liking or preference for a product.
