Sampling & Statistics Flashcards
(81 cards)
1
Q
It refers to how close a particular measure is to the true or correct value
A
Accuracy
2
Q
It is a measure of how reproducible or how close replicate measurements become
A
Precision
3
Q
Errors in Chemical Analysis
A
- Systematic or Determinate error
- Random or Indeterminate error
- Gross error or Blunders
4
Q
Type of error that produces results that constantly deviate from the expected value in one direction or the other
A
Systematic or Determinate error
5
Q
Type of error that is always present in any analytical measurement which is due to our natural limitations in measuring a particular system
A
Random or Indeterminate error
6
Q
Type of error that usually results to a scattered experimental data and not close to the expected value
A
Gross errors or Blunders
7
Q
How can systematic errors be corrected?
A
- Proper calibration of instruments
- Running blank determinations
- Using a different analytical method
8
Q
How can indeterminate errors be corrected?
A
- Essentially unavoidable but it is very small
- Can be minimized by increasing number of measurements
9
Q
How can gross errors be corrected?
A
- Easily identifiable and corrected
- Often leads to outliers
10
Q
3 examples of Systematic error or Determinate
A
- Pipette consistently delivering the wrong volume
- Contaminants in impure reagents, interfering w/ the analysis
- Consistently misreading a meniscus in a burette
11
Q
3 examples of Random error or Indeterminate
A
- Holding the pipette in a different way during measurements
- Uncontrolled changes in temperature or humidity in the lab
- Fluctuations in the readings of instruments
12
Q
3 examples of Gross error or Blunders
A
- Using the wrong reagent or instrument
- Wrong formula in calculations
- Forgetting a step in the analytical procedure
13
Q
It is the lowest possible amount that can be detected w/ some degree of confidence
A
Limit of Detection (LOD)
There is a lower limit at which point we are not sure if something is present or not
14
Q
It relates to the magnitude of change of an instrument w/ changes in compound concentration
A
Sensitivity
It is an indicator of how little can be made before there is a difference on a digital readout
15
Q
The ability of a particular analytical method to detect only the component of interest
A
Specificity
16
Q
It is the lowest amount of analyte which can be quantitatively determined with acceptable precision and accuracy
A
Limit of Quantification (LOQ)
17
Q
Statistical tests that can be used to determine if a suspected data is an outlier
A
-
Q test
Q = |suspect - nearest| / range -
Grubbs test
G = |suspect - mean|/ std dev
Q < Qcrit: retain suspected value
G > Gcrit: reject suspected value
18
Q
Steps in Analysis
A
- Select & prepare sample
- Perform the assay
- Calculate & interpret the results
19
Q
Factors in Method Selection in Food Analysis
A
- Objective of assay
- Characteristic of methods
- Validity of methods
- Consideration of food composition
20
Q
It is the process of proving the method is accurate, reliable, & suitable for its intended purpose
A
Validity of the Method
21
Q
Characteristics of Methods
A
- Inherent properties
- Applicability of the method to the laboratory
- Usefulness
- Personnel
22
Q
Inherent properties of methods
Characteristics of Methods
A
- Specificity/Selectivity
- Precision
- Accuracy
23
Q
Characteristics of Methods related to the Applicability of method to laboratory
Characteristics of Methods
A
- Reagents
- Equipment
- Cost
- Applicability to food/sample
24
Q
Characteristics of methods related to Usefulness
A
- Time required
- Reliability
- Need
25
Characteristics of methods related to Personnel
* Safety
* Procedures
26
It refers to determining the major components of the food matrix
Proximate Analysis of Foods
27
True or False: The performance of many analytical methods is affected by the food matrix
**TRUE**
The major chemical components, especially lipids, proteins, & carbohydrates make up the complex food systems that would require more than technique.
28
True or False: Digestion procedures & extraction methods are not dependent on the food matrix
**FALSE**
Digestion procedures & extraction methods are necessary for accurate analytical results and are dependent on the food matrix
29
It is a well-organized document that establishes the goals of sampling plan, factors to be measured, sampling pount, sampling procedure, frequency, size, personnel, preservation of samples, etc.
Sampling Plan
30
Enumerate 8 different sampling purposes
* Nutritional labeling
* Detection of contaminants & foreign matter
* Acceptance of raw materials, ingredients or products
* Process control samples
* Release of lots of finished products
* Detection of adulterations
* Microbiological safety
* Authenticity of food ingredients
31
Problems in Sampling
Sampling bias which may include
* Selection bias
* Size bias
* Homogeneity bias
32
Problems in Sample storage
* Storage & handling
* Cross-contamination
* Instrumental or preparation
* Mislabeling of samples
33
These are used to describe numeric data from **interval or ratio** measurements, under the assumption of a normal population or distribution
Parametric tests
## Footnote
Parametric statistics require the assumptions of a normal population or distribution
34
These are used to describe **nominal or ordinal** data, wherein the data is not required to fit a normal distribution
Nonparametric Tests
35
Parametric Tests for Independent groups
* Z-test
* Two-sample t-test
* ANOVA
36
Parametric tests for Related groups
Paired t-test
37
Nonparametric tests for Nominal data
* Binomial test
* Chi-squared test
38
Nonparametric tests for Independent groups & Ordinal data
* Mann-Whitney U test
* Kruskal-Wallis test
39
Nonparametric tests for Related groups & Nominal data
Friedman test
40
# **Parametric test & Data**
Parametric test for large sample size (n ≥ 30), std dev is known
Z-test
41
# **Parametric test & Data**
Parametric test for two groups w/ small sample (n < 30), std dev is unknown
Two sample t-test
42
# **Parametric test & Data**
Parametric test for matched samples or related groups
Paired t-test
43
# **Parametric test & Data**
Parametric test for 3 or more groups
Analysis of Variance (ANOVA)
44
# **Nonparametric test & Data**
Nonparametric test for binary outcomes
Binomial test
45
# **Nonparametric test & Data**
Nonparametric test for categorical
Chi-squared test
46
# **Nonparametric test & Data**
Nonparametric tests for ranks from independent groups
Mann-Whitney U test
47
# **Nonparametric test & Data**
Nonparametric test for ranks of 3 or more independent groups
Kruskal-Wallis test
48
# **Nonparametric test & Data**
Nonparametric tests from same panelist
Friedman test
49
Errors in Hypothesis Testing
Type I error
Type II error
50
Error in Hypothesis Testing wherein the null hypothesis (Ho) is rejected when it is true
Type I error
| Ho is true; Reject Ho (False positive)
51
Error in Hypothesis Testing wherein the null hypothesis (Ho) was failed to reject when it is false
Type II error
| Ho is false; Fail to reject Ho (False negative)
52
True or False:
When p-value is < α, reject Ho
**TRUE**
| If p-value is < α, reject Ho
53
True or False:
When p-value is ≥ to α, fail to reject Ho
**TRUE**
| If p-value is ≥ to α, fail to reject Ho
54
This type of bias occurs when the sample is not representative of the population due to the sampling method
Selection bias
55
This type of bias occurs when certain individuals do not respond, and their non-participation is related to the topic being studied
Nonresponse bias
56
This type of bias occurs when the method of observation tends to produce values that systematically differ from the true value in some way, maybe due to leading questions, interviewer influence, or poorly calibrated instruments.
Measurement (response) bias
57
Non-probability sampling methods (4)
* Quota sampling
* Restricted sampling
* Convenience sampling
* Judgement sampling
58
Probability sampling methods (6)
* Simple random sampling
* Cluster sampling
* Probability sampling
* Composite sampling
* Stratified sampling
* Mixed or multi-stage sampling
59
It is a variable that must be avoided since it influences both the response variable and the experimentral conditions
Confounding variables
60
Regression analysis
y = a + bx
a = y-intercept
b = slope
61
# **Sampling method**
A sensory panelist wants 30 participants, 25 male and 25 female, age 18-35. They go to a mall and pick people who fit the criteria until the quota is filled.
Quota sampling (Non-probability)
## Footnote
Quota Sampling: population is grouped into different categories and sample taken from each group until a specific quota is reached.
62
# **Sampling method**
Population is grouped into different categories and sample taken from each group until a specific quota is reached
Quota sampling (Non-probability)
63
# **Sampling method**
A researcher wants to survey only customers who regularly buy organic vegetables from a specific grocery chain.
Restricted sampling (Non-probability)
## Footnote
Restricted sampling - uses some criteria
like geography, brand use, or consumption behavior
64
# **Sampling method**
Sample consists of parts of the population that are easily accessed
Convenience sampling (Non-probability)
65
# **Sampling method**
Sampling is at the discretion of the person collecting samples.
Judgement sampling (Non-probability)
66
# **Sampling method**
Sample is selected randomly from the entire population wherein each unit possess an equal chance of being selected.
Simple Random sampling (Probability)
67
# **Sampling method**
The population is divided into groups called clusters then a subsection is randomly selected
Cluster sampling
68
# **Sampling method**
Sample is formed by selected units at regular intervals starting from a random point.
Systematic sampling
69
# **Sampling method**
Individual samples are combined into one "composite" sample
Composite sampling
70
# **Sampling method**
Subjects are divided into subgroups called strata based on commonality
Stratified sampling
71
# **Sampling method**
It uses some criteria like geography, brand use, or consumption behavior
Restricted sampling (Non-probability)
72
# **Sampling method**
It is a combination of two or more sampling methods; can be random or non-random
Mixed or Multi-stage sampling
73
# **Sampling method**
A food scientist conducting a quick sensory test on a new chocolate variant uses students in the cafeteria at lunch
Convenience sampling (Non-probability)
## Footnote
Convenience sampling - sample consists of parts of the population that are easily accessed
74
# **Sampling method**
A company testing a new high-protein cereals selects athletes or dietitians for a sensory evaluation. These people are not randomly chosen but are judged to be more relevant due to their background and experience.
Judgement sampling (Non-probability)
## Footnote
Judgement sampling - sampling is at the discretion of the person collecting samples.
75
# **Sampling method**
A quality control analyst selects 10 out of 100 packaged juice products from a production batch using a random number generator to test for Brix (sugar content)
Simple Random sampling (Probability)
## Footnote
Simple Random sampling - sample is selected randomly from the entire population wherein each unit possess an equal chance of being selected.
76
# **Sampling method**
A national food chain wants to assess customer satisfaction. They randomly select 10 branches across the Philippines, then survey all customers visiting those branches within a week.
Cluster sampling
## Footnote
Cluster sampling - the population is divided into groups called clusters then a subsection is randomly selected
77
# **Sampling method**
In a production line of bottled sauces, a technician samples every 20th bottle after randomly selecting a starting point between 1 and 20. These samples are checked for pH and viscosity.
Systematic sampling
## Footnote
Systematic sampling - sample is formed by selected units at regular intervals starting from a random point.
78
# **Sampling method**
A milk processor collects 50 ml of milk from 10 randomly selected tanks, mixes them together, and analyzes the composite sample for total bacterial count.
Composite sampling
## Footnote
Composite sampling - individual samples are combined into one "composite" sample
79
# **Sampling method**
For a study on consumer acceptance of a new ice cream flavor, consumers are stratified by age group (teen adults, seniors). A random sample is drawn from each age group to ensure fair representation
Stratified sampling
## Footnote
Stratefied sampling - subjects are divided into subgroups called strata based on commonality
80
# **Sampling method**
In a national survey on food safety practices in restaurants: (1) Randomly select 5 cities (cluster sampling); (2) From each city randomly select 10 restaurants (simple random); (3) From each restaurant, systematically choose 5 staff members to interview.
Mixed or Multi-stage sampling
## Footnote
Mixed sampling - combination of two or more sampling methods; can be random or non-random
81
