Making an Electromagnet by Stephen DeMeo

An electromagnet is created and tested with paperclips. The purpose of this PowerPoint is to provide a template for a rigorous design and analysis of an experimental inquiry. It is primarily for graduate science education students enrolled in my course, 660 Challenging Concepts in Science.

1. THE PHYSICAL EVENT:
MAKING AN ELECTROMAGNET

2. THE PHYSICAL EVENT
I am making an electromagnet by winding copper wire around an
iron nail, attaching the wire ends to a battery, and then seeing
how many paperclips it picks up.

3. MY PRIOR KNOWLEDGE ABOUT THE EVENT
We learned in class that an electromagnet involves electrons and
magnets. So how are these two things related?
When atoms charged with electrons are grouped and move
together, a current is formed. When this current is wound or
coiled around a substance, it can become magnetized.
Electromagnets are all about increasing the magnetic field of a
substance. Because all matter is made up of charged particles
which have a spin, there are magnetic fields for all matter.
Electromagnets help build up the strength of a magnetic field to a
value to show larger scale magnetization in two different ways.

4. MY PRIOR KNOWLEDGE ABOUT THE EVENT
1. The type of substance is important. Iron, cobalt, and
nickel work well because many of the electrons that spin
around the nucleus of these atoms spin in the same
direction, thus producing a larger magnetic field.
2. The wire holding the current must coil the substance
since using a single wire won't make a magnetic field
strong enough to pick up metal objects.
Once the current is turned off, the substance no longer acts
like a magnet.

5. OBSERVING, QUESTIONING, THINKING
ABOUT VARIABLES AND THOSE THAT CAN
BE MEASURED
Observations: With multiple winds of wire around the nail, the nail
becomes magnetized; it attracts paper clips so it does indeed work;
there isn’t any “sensation” when it is close to or when it touches
my skin. The battery gets warm when the wires are connected.
Questions that come to mind: How many winds will pick up the
most paperclips. If I vary the winds of wire will I get a noticeable
difference in how many paperclips can be picked up. What would
happen if I used a bigger battery like a car battery?

6. OBSERVING, QUESTIONING, THINKING
ABOUT VARIABLES AND THOSE THAT CAN
BE MEASURED
Possible Variables to Investigate (M for measurable)
Number of coils of the wire (M), length of the wire (M), number of
paper clips (M), charge on battery (M), size of paper clips (M), type
of metal of the paper clips (?), thickness of the copper wire (M),
weight of the paper clips (M)

7. RESEARCH QUESTION
How does increasing the turns of wire on a
homemade electromagnet (independent
variable) affect the strength of that
magnet (dependent variable)?

8. PREDICTION
The greater the number of coils, the greater
the strength of the magnet

9. 1.Number of Coils of Wire: Independent Variable
2.Number of Paper clips picked up (magnetic strength):
Dependent Variable
3.How tightly the coils are wound (the spaces between
the coils)
4.The charge of the battery
5.The heat generated by the battery
INVENTORY AND CONTROL OF VARIABLES
Inventory of Variables

10. 6.The size and weight of the paperclips
7.The area of the paperclip pile
8.The length of the wound coils
Inventory of Variables
INVENTORY AND CONTROL OF VARIABLES

11. 1.Number of Coils of Wire: Use labels
2.Number of Paper clips picked up (magnetic strength):
Immediately record data in notebook
3.How tightly the coils are wound (the spaces between
the coils): Anchor the spool with a pencil and box; use
tape
4.The charge of the battery: Buy new; don’t wait long
between trials
Controlling these Variables
INVENTORY AND CONTROL OF VARIABLES

12. 5.The heat generated by the battery: Disconnect quickly
after each trial
6.The size and weight of the paperclips: Don’t use large
clips; always use the same ones
7.The area of the pile of paperclips: Try to keep the
same circle size
8.The length of the wound coils: Try to keep them the
same length
Controlling these Variables
INVENTORY AND CONTROL OF VARIABLES

13. I originally thought a 9-volt battery would work well, but a
larger 6-volt lantern battery was much better. I found out that
the 6-volt battery could produce much more electrical current
for a longer time due to its larger electro-chemical cells within
the battery as compared to the smaller 9-volt.
I had old iron nails so I used them, but common nails sold in
hardware stores should work well also.
DESIGN DEVELOPMENT

14. DESIGN DEVELOPMENT: MATERIALS
●Wire
●enamel coated magnet wire 30 AWG
●This smaller gauge facilitates wrapping
the nail with a higher amount of turns
●Battery - 6v lantern battery
●Iron/Steel Nail
●I used 6 iron hand forged nails
●Use nails with as high of an iron content
as you can find! I had iron nails on hand,
but some steel nails will work as well.

15. DESIGN DEVELOPMENT: MATERIALS
●Sandpaper - 220 grit
●Alligator Clips
●Steel Paper Clips
●Tape, Scissors & Marker

16. AGREED TO DESIGN: PROCEDURE
Making the electromagnets
1. The first step is to set up the wire to start
wrapping the nails. It is important to properly set up
the wire so it does not coil on itself.
➢GLP: I suggest creating a spool with a box & pencil
2. Wrap the nail with the wire at one end, leaving a
tail of about 4” of wire.
➢GLP: Keep track of where you started your turns,
ensuring a correct count. Also, try and keep the turns
tight & neat.
Notes for
Good
Laboratory
Practices!

17. 3. Wrap the nail until you have made 50 turns.
➢GLP: If your nail is tapered, like mine, be sure to only
wrap along the end with a consistent width, to ensure each
turn is the same size. It is okay if the wire does not cover
the entire nail, as long as each nail has the same length
covered.
4. When done, trim the wire by cutting a
second 4” tail.
➢GLP: Place a piece of masking tape on the end,
holding the end of the wire to prevent unraveling.
AGREED TO DESIGN: PROCEDURE

18. 5. Lightly rub off the enamel on the ends of the
wire with sandpaper.
➢GLP: Ensure not to rub off too hard/long, the
wire will snap off.
6. Repeat steps 1-5 with two other nails making
one with 100 turns and one with 150 turns.
➢GLP: Be sure to label which is which, so you don’t
mix up each nail.
AGREED TO DESIGN: PROCEDURE

19. Testing the strength in terms of how many
paper clips can be picked up
7. Make a shallow pile of the paperclips on the
table.
8. Holding the nail by the wire ends, lower the
nail straight down into the pile of nails and pull
it back up slowly.
9. Move the nail off to the side and unclip the
nail from the battery, letting the paperclips fall.
➢GLP: Make sure not to mix the piles of paper clips.
AGREED TO DESIGN: PROCEDURE

20. AGREED TO DESIGN: PROCEDURE

21. 10. Count the amount of paper clips that the electromagnet held
and record. Repeat until a pattern emerges and stabilizes.
➢GLP: If you leave the nail connected for too long it will get very hot, so
disconnect it when finished.
11. Remove the tape label during the paperclip test to ensure the
entire length of the nail was available to hold clips.
➢GLP: Promptly replace the labels when you are done to make sure you
do not mix anything up!
12. Repeat for all three nails, record the data in a table.

AGREED TO DESIGN: PROCEDURE

22. COLLECTING RAW DATA: OBSERVATIONS
The wire tips and nail get warm when connected to the battery.

23. COLLECTING RAW DATA: MEASUREMENTS
NUMBER OF PAPERCLIPS HELD PER WIRE TURNS
Wire
Turns Paperclips held per 6 trials
50 8 4 5 8 5 4
100 17 18 20 19 14 21
150 22 24 18 22 26 24

24. FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET
The best representative center value can be
determined by blindly finding the average of the data
set, but a better way is to visualize the data using a
frequency distribution chart.
A fairly symmetric data set is best represented by a
mean, a skewed distribution is best represented by a
median, and a mode is good for a very repetitive
dominant value.

25. FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET
2 2
0 0
2
4 5 6 7 8
Frequency Distribution of Paperclips
picked up by the 50 Turn
Electromagnet
Since the
distribution is
skewed right I
will choose the
median of 5 as
the best
representative
center value

26. This distribution is
skewed left so I
will take the
median of 18 as
the best
representative
center value
1
0 0
1 1 1 1 1
14 15 16 17 18 19 20 21
Frequency Distribution of Paperclips
picked up by the 100 Turn
Electromagnet
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

27. Since this
distribution is
only slightly
skewed, I will
take the average
of 23 as the best
representative
center value
1
0 0 0
2
0
2
0
1
18 19 20 21 22 23 24 25 26
Frequency Distribution of Paperclips
picked up by the 150 Turn
Electromagnet
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

28. Wire
Turns Paperclips held
best
representative
center value of
the Data Set
50 8 4 5 8 5 4 5
100 17 18 20 19 14 21 18
150 22 24 18 22 26 24 23
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

29. PATTERN IDENTIFICATION
A linear pattern is a
possibility
0
5
10
15
20
25
30
0 50 100 150 200
Numbers of Paperclips Held
Turns of Wire on Electromagnet
Numbers of Paperclips Held
vs. Turns of Wire

30. While this data shows a relationship, there are NOT
enough data points to feel confident in making a
conclusion that a linear trend exists.
We need more data!
PATTERN IDENTIFICATION

31. ➢Let’s make 3 more magnets, filling in some
more data points on the trendline:
25, 75, and 125 turns
➢Let’s also add 0 turns and see if the nail picks
up any paperclips. Perhaps the nail is slightly
magnetized.
PATTERN IDENTIFICATION

32. COLLECTING MORE RAW DATA: MEASUREMENTS
Turns Paperclips held
0 0 0 0 0 0 0
25 3 2 3 2 2 3
75 11 12 12 13 1112
125 18 19 20 19 2324

33. Average is 2.5 so the
best representative
center value could
be 2 or 3. I will use
my rounding rules
and choose 2
0
3 3
0
1 2 3 4
Frequency Distribution of Paperclips
picked up by the 25 Turn
Electromagnet
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

34. Since distribution is
fairly bell-shaped I
will take the average
of 12 as my best
representative
center value
0
2
3
1
0
10 11 12 13 14
Frequency Distribution of
Paperclips picked up by the 75
Turn Electromagnet
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

35. Since the
distribution is
skewed right I will
use the median of 20
as the best
representative center
value
1
2
1
0 0
1 1
18 19 20 21 22 23 24
Frequency Distribution of
Paperclips picked up by the 125
Turn Electromagnet
FINDING THE BEST REPRESENTATIVE CENTER
VALUE FROM THE DATA SET

36. •In terms of center, most best representative center values
were medians.
•In terms of shape, most of the frequency plots were skewed.
•In terms of spread, some of the diagrams were more spread
out than others.
•In terms of unusual features there were some gaps between
the bins.
Features of the Frequency Distribution Diagrams

37. COLLECTING ALL RAW DATA: MEASUREMENTS
Wire
Turns Paperclips held
Best Rep.
Value of the
Data Sets
0 0 0 0 0 0 0 0
25 3 2 3 2 2 3 2
50 8 4 5 8 5 4 5
75 11 12 12 13 11 12 12
100 17 18 20 19 14 21 18
125 18 19 20 19 23 24 20
150 22 24 18 22 26 24 23

38. PATTERN IDENTIFICATION
0
5
10
15
20
25
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

39. PATTERN IDENTIFICATION
With Excel’s trendline function, I will try to find the best
trendline, linear or curved, that fits the data (bisects the
data points)

40. PATTERN IDENTIFICATION
Power
A power
trendline
doesn’t fit the
data so well.
(zero point
can’t be
included)
y = 0.0205x
1.4346
0
5
10
15
20
25
30
0 20 40 60 80 100 120 140 160
Numbers of Paperclips Held
Turns of Wire on Electromagnet
Numbers of Paperclips Held vs. Turns of Wire

41. PATTERN IDENTIFICATION
Logarithmic
A logarithmic
trendline also
doesn’t fit the
data so well.
(zero point
can’t be
included)
y = 12.425ln(x) - 40.285
-5
0
5
10
15
20
25
0 20 40 60 80 100 120 140 160
Numbers of Paperclips Held
Turns of Wire on Electromagnet
Numbers of Paperclips Held vs. Turns of Wire

42. PATTERN IDENTIFICATION
Exponential
An exponential
trendline also
doesn’t fit the
data so well.
(zero point
cannot be
included)
y = 1.8662e
0.0192x
0
5
10
15
20
25
30
35
0 20 40 60 80 100 120 140 160
Numbers of Paperclips Held
Turns of Wire on Electromagnet
Numbers of Paperclips Held vs. Turns of Wire

43. PATTERN IDENTIFICATION
Polynomial
When the zero
point is
included, the
polynomial
trendline fits
the data better
and resembles
a straight line
This is because
the x
2
term is
very small.
y = -4E-05x
2
+ 0.1743x - 1.3333
-5
0
5
10
15
20
25
30
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

44. PATTERN IDENTIFICATION
Linear
A straight line
or linear
relationship
also fits the
data set well
(zero included)
y = 0.1686x - 1.2143
-5
0
5
10
15
20
25
30
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

45. I will choose the linear plot over the polynomial without the zero
since more data points rest on the line. When the zero point is
included, the straight line and polynomial trendlines are almost
identical.
PATTERN MEANING

46. ➢We can describe the plot’s characteristics in terms of shape,
slope, strength and unusual features.
Shape:
Excel provides an equation for our linear trendline:
y = 0.1686x – 1.2143
Substituting our variables and rounding, our linear mathematical
model is:
Number of Paperclips Held
=
(0.17 paperclips/turns) (Wire Turns) – 1 paperclips
PATTERN MEANING

47. Slope and y-intercept:
Our equation says our slope is 0.17 paperclips held per 1 turn of
wire. Our slope means that an electromagnet consisting of a single
turn of wire should be able to hold 0.17 paper clips; more
realistically, it means that a 100 turn electromagnet should hold
17 paper clips. Incorporating the y-intercept would give us:
17 – 1 = 16 paperclips.
Strength: The linear relationship looks fair to good with some points
somewhat close to the line.
Unusual Features: none were observed
PATTERN MEANING

48. ➢Because the pattern of the data strongly suggests a straight
line pattern from 0 to 150 turns, It means that our two
variables, paperclips held and turns of wire, closely predict
each other in a specific proportional manner.
➢Keeping the current of the battery constant, the turns of the
wire produces an electromagnet, with greater turns
producing greater magnetic strength.
PATTERN MEANING

49. PRELIMINARY ANSWER TO THE
RESEARCH QUESTION
Research Question:
How does increasing the turns of wire on a homemade
electromagnet affect the strength of that magnet?
As the amount of turns of wire increases, the number of
paper clips able to be held also increases in a
proportional, linear manner represented by the equation:
y = 0.17 paperclips held/turn – 1 paperclip, where y is
the number of paperclips held.

50. CONFIDENCE INDICATORS

51. Good (and Ethical) Laboratory Practices
(emphasizes Reliability and Validity)
➢We ensured GLP by following the protocol carefully and
mindfully. GLP notes were recorded in the procedure.
➢Some important GLPs involved being consistent from
one trial to the next in relation to the winding of the
coils and the dispersion of the paper clips.

52. Repetition (emphasizes Reliability)
➢For each condition we repeated the trials until:
•A pattern emerged
•Was reasonable with materials and time
•Was safe and careful (so the battery didn’t burn
out).
Peer Replication
➢Lily O’Heir’s Data
Peer Review (emphasizes Validity)
➢See the slide near the end of this presentation

53. Wire
Turns
Paperclips held per 6 trials Best Rep. Value
of the Data Set
0 0 0 0 0 0 0 0
25 18 14 14 15 13 16 15
50 21 18 27 22 28 24 23
75 28 39 28 31 33 23 30
100 29 30 37 25 32 42 32
125 37 28 32 40 29 41 34
150 42 38 43 29 34 35 37
Peer Replication: Lily O’Heir (emphasizes Reliability)

54. Peer Replication: Pattern Identification
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

55. Peer Replication: Pattern Identification
Power
A power
trendline seems
to fit the data
fairly well (zero
point cannot be
included)
y = 3.1702x
0.4992
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

56. Peer Replication: Pattern Identification
Logarithmic
A log trendline
also seems to
fit the data
fairly well (zero
point cannot be
included)
y = 12.193ln(x) - 24.118
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

57. Peer Replication: Pattern Identification
Exponential
An exponential
trendline does
not fit the data
(zero point
cannot be
included)
y = 15.376e
0.0066x
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140 160
Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

58. Peer Replication: Pattern Identification
Polynomial
A polynomial (2
nd

order quadratic)
trendline with the
zero point
included also fits
the data well
y = -0.0019x
2
+ 0.5114x + 1.5476
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

59. Peer Replication: Pattern Identification
Linear
A linear
trendline is not
a good fit.
y = 0.2257x + 7.5
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140 160Number of Paperclips Held
Turns of Wire
Paperclips Held vs. Turns of Wire

60. Peer Replication
The logarithmic and polynomial (2
nd
order quadratic) plots
seems to fit the data slightly better than the power, and
much better than the linear and exponential trendlines.
Therefore in terms of a mathematical equation, this peer
data does not support the main experiment’s linear
relationship. The reason is not known at this time.
What is common between the main and this peer experiment
is the positive qualitative trend of the data. In both, the
amount of paperclips increased as the number of turns of
wire increased.

61. Comparison with Material Standard
(emphasizes Validity)
➢This indicator was not possible since no standard was available.

62. Secondary Test (emphasizes Reliability)
➢I created a secondary way of testing
the strength of our magnets by using
staples instead of paperclips.
➢ I followed the same exact procedure as
the paper clip test.
GLP: The staples need to be laid completely
flat, because they link on each other easily.

63. Secondary Test
➢Observations:
The electromagnet picked up large bunches of staples in a
similar way to the paperclips.

64. Secondary Test: Raw Data
Turns Staples held Best Rep.
Value
0 0 0 0 0 0 0 0
25 4 5 5 4 4 3 4
50 9 5 7 5 4 6 6
75 13 15 14 14 17 15 15
100 16 15 17 17 18 18 17
125 26 25 30 35 26 27 28
150 40 32 37 45 38 32 37

65. Secondary Test: Pattern Identification
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

66. Secondary Test: Pattern Identification
Power
A power
trendline fits
the data
better (zero
point could
not be
included)
y = 0.0551x
1.2767
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

67. Secondary Test: Pattern Identification
Logrithmic
A log
trendline isn’t
very good
(zero point
could not be
included)
y = 17.415ln(x) - 57.32
-5
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

68. Secondary Test: Pattern Identification
Exponential
An exponential
trendline is a bit
better than the
power trendline:
more data
points on the
line (zero point
could not be
included).
y = 2.7828e
0.0181x
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

69. Secondary Test: Pattern Identification
Polynomial
A quadratic
(2
nd
order
polynomial)
trendline is
much better.
y = 0.0011x
2
+ 0.0829x + 0.4048
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

70. Secondary Test: Pattern Identification
Linear
A linear
trendline
doesn’t fit
the data very
well
y = 0.2429x - 2.9286
-5
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160
Number of Staples Held
Turns of Wire
Staples held vs. Turns of Wire

71. The exponential and polynomial trendlines fit the data best.
This secondary test contradicts the main experiment (linear
trendline was best), and to some extent the replicated
experiment (polynomial and logarithmic trendlines).
Secondary Test: Pattern Identification

72. Plots and Patterns (emphasizes Reliability)
Data that follows a specific pattern provides greater
confidence that the experiment is being conducted correctly.
➢In the main experiment, three more data points and a zero
trial were added to our existing data set which confirmed
the initial pattern in the data.
➢An overall increasing qualitative trend between the
variables was evident in the main, peer experiment, and in
the secondary test.

73. Conceptual & Historical Support (emphasizes Validity)
Only the main experiment is supported by the formula for
“Magnetic Field Strength” or “Magnetizing Force” represented by H.
H = I x N
L
Where:
N is the number of turns of the coil
I is the current flowing through the coil in amps from the battery
L is the length of the coil

74. Conceptual & Historical Support
The equation indicates that the greater the number of coils (N),
the greater the magnetic field strength (H). There is a direct
positive proportional relationship between N and H, if L and I are
constant. This relationship will give a linear plot.
This provides confidence for the linear relationship found in the
main experiment.

75. Conceptual & Historical Support
The first electromagnet was created by
William Sturgeon of England in 1825. It was a
horseshoe-shaped piece of iron that had 18
turns of bare copper wire wrapped around it.
Sturgeon used varnish to insulate the iron
metal from the copper coil (insulated wire
didn't exist yet).
An electromagnet
made in the 1800’s

76. External Data Sources (emphasizes Reliability)

I did not find any data sources that would support or
challenge my data.

77. Precision (emphasizes Reliability)
Instrumental Limit of Precision
➢There are no measuring instruments used in this experiment.
Counting is not measurement.

78. Precision
Data Set Precision
➢Data Set Precision can be measured with the average
deviation statistic. This can tell us the variability or spread of
the data set for each set of trials.

79. Precision: Data Set Precision
Wire
Turns Paperclips held
Average
Deviation
(+/–)
Average
Deviation in
(+/–) Percent
0 0 0 0 0 0 0 0.0 0%
25 3 2 3 2 2 3 0.5 25%
50 8 4 5 8 5 4 1.6 32%
75 11 12 12 13 1112 0.5 4%
10017 18 20 19 1421 1.8 10%
12518 19 20 19 2324 1.8 9%
15022 24 18 22 2624 2.0 9%
Average = +1.2 +13%

80. Precision: Data Set Precision
➢The smallest average deviation was +0.5 while the largest
was +2.0 paperclips.
➢In terms of percent average deviation, the smallest was 4%
and the largest was 32% which is a bit high mainly because
of the low number of paperclips picked up by the magnet.
➢The average of the average deviations was about +1
paperclip, and the average of the percent average deviation
was +13%. These values exhibited acceptable precision
especially considering the crude materials that were used.

81. Accuracy (emphasizes Validity)
Mean Absolute percent accuracy for trendlines
➢To use this statistic, I will have to collect Y observed
values using my actual data points. Y predicted values
will be collected by using the equation of the trendline
calculated by Excel: y=0.1686x–1.2143. Substituting in
the corresponding x values will give me my Y predicted
values.

82. Accuracy: mean absolute accuracy

X Value
(Wire Turns)
Yo Observed
(Paperclips)
Yp Predicted from
Trendline
(Paperclips)
0 0 -1.2143
25 2 3.0007
50 5 7.2157
75 12 11.4307
100 18 15.6457
125 20 19.8607
150 23 24.0757

83. Accuracy: mean absolute percent accuracy

There is a problem! When I use the online calculator
(search for “MAPE Calculator”), and enter my Y observed
and predicted values, I get an infinity result. This is
because there is a zero value for Y observed.
The workaround is to get rid of that data (the first row)
or not to calculate percent and use a mean absolute
accuracy statistic instead. I chose the latter.
Using a MAE calculator, the mean absolute accuracy is
+1.2 paperclips

84. Accuracy: mean absolute accuracy

What does the mean absolute accuracy mean?
➢This statistic measures how far off the data points are
to a linear trendline (the bisecting line aka the
predicted trend that Excel has made).
➢To understand this better, visualize a straight line that is about
+1 paperclip unit parallel and above the trendline in the plot
as well as almost –1 paperclip unit parallel and below the
trendline. This zone represents how far off overall the data
points are to the trendline.

85. Accuracy: mean absolute accuracy
➢The mean absolute accuracy of +1.2 demonstrates a good
level of accuracy for this experiment.

86. Redesign & What If Ideas
(emphasizes Validity & Reliability))
➢Redesign: A redesign was made with the inclusion of three
more data points plus a zero point in order to determine if a
linear or curved relationship was present.
➢What If Idea: As a future secondary test, the electromagnets
could be placed near a compass and the degree to which the
needle moves away from true north can be measured.
Check Calculations (emphasizes Reliability)
➢The frequency distribution calculations were checked.

87. Confidence
Indicator
Relevant to
this
Experiment
Directly Verified
Preliminary
Result
Contradicted
Preliminary Result
or Inconclusive
Good Lab Practices ✓
Repetition ✓ ✓
Peer Support ✓ ✓
Secondary Test ✓ ✓
Comparison with a
Material Standard
Plots and Patterns ✓
Conceptual & Historical
Support
✓
External Data Sources
Precision ✓
Accuracy ✓
Check Calculations ✓
Redesign & What If Ideas ✓
SUMMATIVE ANALYSIS OF THE CONFIDENCE
INDICATORS
The following
table shows
which of the
12 indicators
were used and
which ones
verified or did
not verify the
preliminary
result.

88. CONFIDENT ANSWER TO THE RESEARCH
QUESTION
➢I have low confidence that a linear relationship exists in my
main experiment. This experiment could not be reproduced by
a peer and a secondary test was not supportive.
Research Question:
How does increasing the turns of wire on a homemade
electromagnet affect the strength of that magnet?

89. CONFIDENT ANSWER TO THE RESEARCH
QUESTION
➢I can confidently say that the data definitely supports a
positive relationship between the coil size of the
electromagnet and its strength as measured by the number
of paper clips that could be picked up, but what exactly that
quantitative relationship is, could not be determined by this
experiment.
Research Question:
How does increasing the turns of wire on a homemade
electromagnet affect the strength of that magnet?

90. CONFIDENT ANSWER TO THE RESEARCH
QUESTION
➢My prediction was correct, the more turns of wire the more
paperclips were picked up by the electromagnet.
Research Question:
How does increasing the turns of wire on a homemade
electromagnet affect the strength of that magnet?

91. INVESTIGATOR’S CONFIDENCE LEVEL IN THEIR
ANSWER
In a scale of 1 to 10, I give my confident result a 10 out of 10
that a positive qualitative relationship exists between the coil
size of the electromagnet and its strength as measured by the
number of paper clips that could be picked up, but what
exactly the relationship is, cannot be gleamed from this
experiment.

92. SIGNIFICANCE OF THE ANSWER IN LARGER
CONCEPTUAL AND SOCIAL CONTEXTS
We use electromagnets everyday. They are found in audio
speakers, electric powered saws, beverage mixers, and
hairdryers. And the future of transportation – electric cars –
rely on motors that are electromagnets.

93. The photo of the earth to the right shows
a reddish outer ring or core. This outer
core is made of liquid or molten iron, the
same element the nails in our experiment
are made out of.
The liquid iron of the outer core is so
dynamic, that the constant movement of
iron atoms creates a magnetic field. This
field protects us from damaging radiation
from the sun.
SIGNIFICANCE OF THE ANSWER IN LARGER
CONCEPTUAL AND SOCIAL CONTEXTS

94. “The ancient Greeks and Chinese knew about naturally
magnetic stones called ‘lodestones’. These chunks of iron-rich
minerals may have been magnetized by lightning. The Chinese
discovered that they could make a needle magnetic by stroking
it against a lodestone, and that the needle would point north-
south.”
“Some animals, such as pigeons, bees, and salmon, can detect
the Earth's magnetic field and use it to navigate. Scientists
aren't sure how they do this, but these creatures seem to have
magnetic material in their bodies that acts like a compass.”
SIGNIFICANCE OF THE ANSWER IN LARGER
CONCEPTUAL AND SOCIAL CONTEXTS

95. TITLE
Determining The Strength of an ElectroMagnet

96. SUMMARY (ABSTRACT)
An experiment was performed to determine if the number
of coils around a rudimentary electromagnet would
increase its strength. Strength was assessed by the
number of metal paperclips that could be picked up. It was
determined that as the number of coils increased by
increments of 25 turns, the electromagnet picked up ever
increasing numbers of paperclips. This finding was
confirmed by a secondary test using metal staples and
replicated by a peer. While this qualitative trend was
apparent, a quantitative relationship was not confidently
determined.

97. NAMES AND ROLES
Stephen DeMeo & Ellen Stockbridge – Investigators
Lilly O’Heir – Student who reproduced the experiment

98. ACKNOWLEDGEMENTS
This activity would not have possible if it wasn’t for Ellen
Stockbridge, an exceptional student of mine who is now teaching
earth science students in New York City.

99. REFERENCES
National Geographic Resource Library, Magnetism.
<https://www.nationalgeographic.org/encyclopedia/magnet
ism/>
Stack Exchange, Physics, “Do all things have a magnetic
field”. See
<https://physics.stackexchange.com/questions/187333/do -
all-the-things-have-a-magnetic-field/187338>
Wikipedia, “Electromagnet”.
<https://en.wikipedia.org/wiki/Electromagnet>
“The Electromagnet”. https://www.electronics-
tutorials.ws/electromagnetism/electromagnets.html

100. ✓
Peer Review (emphasizes Validity)
Consensus was reached because a strong majority of
the class agreed with my or our confident result and
the parts that led to its construction.
Consensus was not reached because a majority of the
class disagreed with my or our confident result and
the parts that led to its construction.
Recommendations for Revision: